Citation
Moral Panics and the Sex Offender Registry Are the Laws Really Protecting Communities

Material Information

Title:
Moral Panics and the Sex Offender Registry Are the Laws Really Protecting Communities
Creator:
Klein, Jennifer L
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (350 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Criminology, Law, and Society
Sociology and Criminology & Law
Committee Chair:
LANZA KADUCE,LONN M
Committee Co-Chair:
AKERS,RONALD L
Committee Members:
LEVETT,LORA M
WEBSTER,GREGORY DANIEL
Graduation Date:
5/3/2014

Subjects

Subjects / Keywords:
Correlations ( jstor )
Criminal offenses ( jstor )
Criminals ( jstor )
Modeling ( jstor )
Morality ( jstor )
Panic ( jstor )
Parents ( jstor )
Predators ( jstor )
Registry ( jstor )
Sex offenders ( jstor )
Sociology and Criminology & Law -- Dissertations, Academic -- UF
national -- offender -- perceptions -- registry -- sex -- survey
City of Gainesville ( local )
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Criminology, Law, and Society thesis, Ph.D.

Notes

Abstract:
There is a unique stigma associated with sexual offenders living in the community. Once their sentence is completed, sexual offenders are required to register with the state and to have their personal information made available on a public website. Community members are wary of these perceived predators as threats to their children and as a possible disruption to the innocent lives their children lead. Not all sex crimes are the same but all sex offenders are commonly treated and perceived as the same, no matter their crime. Many individuals (the general public, politicians, etc.) have one perception of sexual offenders, which might not match the reality of their crimes. The literature surrounding community perceptions of sexual offenders is limited by continuously growing. The current study examines the perceptions of a national sample of participants who have children and who do not. This direct comparison of groups adds to the literature by bringing the protective nature of parenthood into play. Previous research has shown that sex offenders do not have high rates of reoffending, but there is fear and stigma associated with these individuals. This dichotomy, between actual rates of recidivism and public fear of reoffending, will serve as the focal point of the paper, in trying to find the cause of these misconceptions. When dealing with social issues such as perceptions and stigma, many factors could potentially influence an individual's perceptions, including their status as a parent, their socio-economic status, their previous experiences with sex offenses and other crimes and exposure to the media. Furthermore, this project will examine the subtlety present in the comparison of sex offenders and sex predators. A random assignment of those terms will be applied to the study to examine which terms is more salient to participants, and whether or not they are able to distinguish the difference in the two terms. All of these factors will be tested in conjunction with registry knowledge, moral panic measures, legal support and accuracy to determine just how important and concerning the presence of sexual offenders is in the eyes of the general citizenry. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (Ph.D.)--University of Florida, 2014.
Local:
Adviser: LANZA KADUCE,LONN M.
Local:
Co-adviser: AKERS,RONALD L.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2016-05-31
Statement of Responsibility:
by Jennifer L Klein.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Applicable rights reserved.
Embargo Date:
5/31/2016
Resource Identifier:
907379383 ( OCLC )
Classification:
LD1780 2014 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

MORAL PANICS AND THE SEX OFFENDER REGISTRY: ARE THE LAWS REALLY PROTECTING COMMUNITIES? By JENNIFER L. KLEIN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2014 1

PAGE 2

2014 Jennifer L. Klein 2

PAGE 3

To Beverly and Albert Klein, the most supportive parents anyone could ask for 3

PAGE 4

ACKNOWLEDGMENTS Although these acknowledgments are significantly premature and woefully incomplete, I would like to thank my committee members, Drs. Lora Levett, Ron Akers and Greg Webster, in advance for taking the time to read this document and for providing constructive criticism that will ultimate make this project better. Specifically, I must thank my chair Dr. Lonn Lanza Kaduce for his dedication in reading this proposal over and over again, and maki ng this study so much better then I could on my own. I also have to acknowledge my research partner, Danielle Tolson, who was integral to keeping me sane throughout this process. Without her, I would have had a much harder time trying to keep everything on track for a defense and graduation that fit within the Klein/Tolson timeline. Thank you. 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ ........ 12 ABSTRACT ................................ ................................ ................................ ................... 13 CHAPTER 1 INTRODUCTI ON ................................ ................................ ................................ .... 15 2 LITERATURE REVIEW ................................ ................................ .......................... 22 Moral Panics Literature ................................ ................................ ........................... 22 Conn ecting Moral Panics and the Sex Offender Registry ................................ ....... 29 High Profile Abduction Cases Lead to New Laws ................................ ................... 35 Community Perceptions of Sexual Offenders ................................ ......................... 40 Community Perceptions of the Sex Offender Registry ................................ ............ 47 Community Attitudes Towards Sexual Offenders (CATSO) Scale .......................... 52 Fear of Sexual Victimization ................................ ................................ ................... 56 Prior Literature on the Stereotypical Sex Offender ................................ ................. 58 3 RESEARCH METHODOLOGY ................................ ................................ ............... 62 Explor atory Hypotheses ................................ ................................ .......................... 66 Exploratory Hypotheses Concerning the "Personal Orientations Toward the Control of Sexual Offenders." ................................ ................................ ........ 67 Exploratory Hypothesis About Fear of Crime (No Randomized Offender/Predator Manipulation) ................................ ................................ ... 67 Exploratory Hypotheses about "Perceptions of Community Related Attitudes" Citizen endorsement of Moral Panic Features (Randomized Offender/Predator Manipulati on) ................................ ................................ ... 68 Participants ................................ ................................ ................................ ............. 69 Recruitment of Participants ................................ ................................ ..................... 70 Pilot Participants ................................ ................................ ............................... 70 Full Study Participants ................................ ................................ ...................... 70 Inst rumentation Pilot Instrument ................................ ................................ ........... 71 Full Study Operationalization ................................ ................................ .................. 75 Full Study Instrument ................................ ................................ .............................. 75 Personal Orientations toward the Control of Sex Offenders ............................. 76 Registry Knowledge ................................ ................................ ......................... 77 Stereotypical Sex Offender ................................ ................................ ............... 78 Fear of Sexual Victimization ................................ ................................ ............. 80 5

PAGE 6

Perceptions of Community Related Attitudes ................................ ................... 81 Concern ................................ ................................ ................................ ..... 81 Hostility ................................ ................................ ................................ ...... 81 Consensus ................................ ................................ ................................ 82 Disproportionality ................................ ................................ ....................... 83 Volatility ................................ ................................ ................................ ...... 85 Community Attitudes Toward Sex Offenders (CATSO) Scale ................... 86 Registry Support ................................ ................................ ............................... 87 Demographic and Control Variables ................................ ................................ ....... 88 Data Analysis Plan ................................ ................................ ................................ .. 90 4 RESULTS ................................ ................................ ................................ ............. 106 Participant Demogr aphics ................................ ................................ ..................... 106 Hypothesis Testing and Results ................................ ................................ ........... 111 Hypothesis Testing Concerning Personal Orientations Toward the Control of Sex Offenders Models ................................ ................................ ....................... 112 Model 1: Across Group Comparison for Sex Offender Registry Website Access ................................ ................................ ................................ ......... 112 Model 2: Registry Knowledge Across Participants and Predicting the Stereotypical Sex Offender ................................ ................................ ......... 114 Predicting the Stereotypical Sex Offender ................................ ...................... 119 Hypothesis Testing for Fear of Sexual Victimization Models ................................ 124 Model 3: OLS Regression Model Predicting Fear of Victimization ................. 124 Hypothesis Testing for Perceptions of Community Related Attitudes Models ...... 131 Factor Analysis for Moral Panic Measures ................................ ..................... 131 Model 4: OLS Regression Model Predicting the Moral Panic Element of Concern ................................ ................................ ................................ ....... 134 Model 5: OLS Regression Predicting the Moral Panic Element of Hostility .... 141 Model 6: OLS Regression Model Predicting the Moral Panic Element of Consensus ................................ ................................ ................................ .. 148 Model 7: OLS Regression Model Predicting the Moral Panic Element of Disproportionality ................................ ................................ ........................ 155 Model 8: OLS Regression Model Predicting the Moral Panic Element of Volatility ................................ ................................ ................................ ....... 163 Model 9: OLS Regression Model Predicting the Community Attitudes Toward Sex Offenders (CATSO) Scale ................................ ....................... 169 Model 10: OLS Regression Model Predicting Registry Support ..................... 184 5 DISCUSSION AND CONCLUSIONS ................................ ................................ .... 208 Discussion ................................ ................................ ................................ ............ 208 Contributions to the Field ................................ ................................ ...................... 211 Study Methodology ................................ ................................ ......................... 211 Theoretical Contributions ................................ ................................ ................ 213 Co nceptualization ................................ ................................ ........................... 216 Discussion of Specific Issues ................................ ................................ ......... 218 6

PAGE 7

Registry Knowledge as an Independent Variable ................................ ........... 218 The Role of the Media as Related to Moral Panics ................................ ........ 221 Younger Participants Mean a Different Knowledge Base ............................... 222 Moral Panics as a Theoretical Lens ................................ ............................... 223 Offender vs. Predator Distinction ................................ ................................ .... 226 Limitations ................................ ................................ ................................ ............. 228 Future Research ................................ ................................ ................................ ... 231 Policy Implications ................................ ................................ ................................ 233 APPENDIX A PILOT STUDY INSTRUMENT ................................ ................................ .............. 239 B PILOT STUDY CODING SHEET ................................ ................................ .......... 247 C PILOT STUDY FREQUENCY AND REGRESSION TABLES ............................... 255 LIST OF REFERENCES ................................ ................................ ............................. 339 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 350 7

PAGE 8

LIST OF TABLES Table page 1 1 Finalized CATSO Scale ................................ ................................ ...................... 54 3 1 Factor Analysis for Registry Knowledge ................................ ............................. 72 3 2 Factor Analysis for Moral Panic Measures ................................ ......................... 74 3 3 Model 1: Across Group Comparison for Registry Website Access ..................... 92 3 4 Model 2: Registry Knowledge Across Participants ................................ ............. 93 3 5 Model 2: Predicting the Stereotypical Sex Offender. ................................ .......... 94 3 6 Model 3: OLS Regression Model Predicting Fear of Victimization ..................... 96 3 7 Model 4: OLS Regression M odel Predicting the Moral Panic Element of Concern ................................ ................................ ................................ .............. 97 3 8 Model 5: OLS Regression Model Predicting the Moral Panic Element of Hostility ................................ ................................ ................................ ............... 98 3 9 Model 6: OLS Regression Model Predicting the Moral Panic Element of Consensus ................................ ................................ ................................ .......... 99 3 10 Model 7: OLS Regression Model Predicting the Moral Panic Element of Disproportionality ................................ ................................ .............................. 100 3 11 Model 8: OLS Regression Model Predicting the Moral Panic Element of Volatility ................................ ................................ ................................ ............ 101 3 12 Model 9: OLS Regression Model Predicting the Community Attitudes Towards Sex Offenders (CATSO) Scale ................................ .......................... 102 3 13 Model 10: OLS Regression Model Predicting Registry Support ....................... 104 4 1 Univariate Analysis for Participant Demographics: Gender, Age, Race and Ethnicity ................................ ................................ ................................ ............ 106 4 2 Univariate Analysis for Participant Demographics: Parental Status, Information Regarding Children, and Grandparental Status ............................. 107 4 3 Univariate Analysis for Participant Demographics: Education, Marital Status, Income Level, Socio Economic Status ................................ ............................. 109 4 4 Univariate Analysis for Participant Demographics: State of Residence, Type of Residence, Population Size, Geographic Region ................................ ......... 110 8

PAGE 9

4 5 Univariate Statist ics for Registry Website Access ................................ ............ 112 4 6 Independent Sample T Test for Parental Status and Gender for Registry Access ................................ ................................ ................................ .............. 114 4 7 Univariate Statistics for Registry Knowledge Across Participants ..................... 115 4 8 Univariate Analysis for Participants' Count Scores for Registry Knowledge ..... 116 4 9 Independent Sample T Test for Parental Status or Gender and Additive Registry Knowledge ................................ ................................ .......................... 119 4 10 Univariate Analysis for the Stereotypical Sex Offender Measures. ................... 120 4 11 Univariate Analysis for the Variable Index for the Ster eotypical Sex Offender Measures. ................................ ................................ ................................ ......... 122 4 12 Independent Sample T Test for Parental Status or Gender and the Stereotypical Sex Offen der Measures. ................................ ............................. 124 4 13 Univariate for Fear of Victimization Responses. ................................ ............... 125 4 14 Independent Sample T Test for Parental Status or Gender and Fear of Victimization ................................ ................................ ................................ ..... 128 4 15 OLS Regression Predicting Fear of Children Being Victimized and Fear of Personal Victimization. ................................ ................................ ..................... 131 4 16 Factor Analysis for Moral Panic Measures. ................................ ...................... 132 4 17 Univariate Analysis for Moral Panic Element of Concern (Offender and Predator Randomization). ................................ ................................ ................. 135 4 18 Independent Sample T Test for Parental Status and the Element o f Concern 138 4 19 OLS Regression Predicting Element of Concern (Random Assignment). ........ 140 4 20 Univariate Analysis for Moral Panic Element of Hostility (Offender and Predator Randomization). ................................ ................................ ................. 142 4 21 Independent Sample T Test for Parental Status and the Element of Hostility .. 146 4 22 OLS Regression Predicting Element of Hostility (Random Assignment). ......... 147 4 23 Univariate Analysis for Moral Panic Element of Consensus (Offender and Predator Randomization). ................................ ................................ ................. 149 4 24 Independent Sample T Test for Parental Status and the Element of Concern 153 9

PAGE 10

4 25 OLS Regression Predicting Element of Consensus (Random Assignment). .... 155 4 26 Univariate Analysis for Moral Panic Element of Disproportionality ( Offender and Predator Randomization). ................................ ................................ .......... 157 4 27 Independent Sample T Test for Parental Status and the Element of Disproportionality ................................ ................................ ............................. 161 4 28 OLS Regression Predicting Element of Disproportionality (Random Assignment). ................................ ................................ ................................ ..... 162 4 29 Univariate Analysis for Moral Panic Element of Volatility (Offender and Predator Randomization). ................................ ................................ ................. 164 4 30 Independent Sample T Test for Parental Status and the Element of Volatility 167 4 31 OLS Regression Predicting Element of Volatility (Random Assignment). ........ 169 4 32 Univariate Analysis for the Community Attitudes Toward Sex Offenders (CATSO) Scale. ................................ ................................ ................................ 171 4 33 Independent Sample T Tests for Parental Status and the Four Constructs of the CATSO Scale ................................ ................................ ............................. 178 4 34 OLS Regression Predicting Social Isolation and Capacity to Change Constructs (CATSO Scale). ................................ ................................ .............. 180 4 35 OLS Regression Predicting OLS Regression Predicting Severity/Dangerousness and Deviancy Constructs (CATSO Scale). ............... 182 4 36 OLS Regression Predicting Total Index of Negative Attitudes (CATSO Scale). ................................ ................................ ................................ .............. 184 4 37 Univariate Analysis for Registry Support (Offender and Predator Randomization). ................................ ................................ ............................... 186 4 38 Independent Sample T Test for Parental Status and the Registry Support (Offender and Predator Randomization). ................................ .......................... 193 4 39 Univariate Analysis for Perceived Level of Strictness. ................................ ...... 194 4 40 OLS Regression Predicting Registry Support (Random Assignment). ............. 195 4 41 Analysis Summary for Personal Orientations toward the Control of Sexual Offenders. ................................ ................................ ................................ ......... 197 4 42 Analysis Summary for Fear of Victimization ................................ ..................... 199 4 43 Analysis Summary for Moral Panic Regressions ................................ .............. 200 10

PAGE 11

4 44 Analysis Summary for Community Related Attitudes. ................................ ...... 205 11

PAGE 12

L IST OF FIGURES Figure page 4 1 Additive Responses for Registry Knowledge Measures ................................ ... 117 4 2 Additive Responses for the Stereotypical Sex Offender Measures. ................. 122 12

PAGE 13

ABSTRACT OF DISSERTA TION PRESENTED TO TH E GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLME NT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY MORAL PANICS AND THE SEX OFFENDER REGISTR Y: ARE THE LAWS REALLY PROTECTING COMMUNITI ES? By Jennifer L. Klein May 2014 Chair: Lonn Lanza Kaduce Major: Criminology, Law and Society There is a unique stigma associated with sex offenders living in the community. Once their sentence is completed, sex offenders are required to register with the state and to have their personal information made available on a public website. Community members are wary of these perceived predators as threats to their children and as a possible disruption to the innocent lives their children lead. Not all sex crimes are the same but all sex offenders are commonly treated and perceived as the same, no matter their crime. Many individuals (the general public, politicians, etc.) have one perception of sex offenders, which might not match the reality of their crimes. The literature surrounding community perceptions of sex offenders is limited by continuously growing. The current study examines the perceptions of a national sample of participants who have children and who do not. This direct comparison of groups adds to the literature by bringing the protective nature of parenthood into play. Previous research has shown that sex offenders do not have high rates of reoffending, but there is fear and stigma associated with th ese individuals. This dichotomy, between actual rates of recidivism and public fear of reoffending, will serve as the focal point of the paper, in trying to find the cause of these misconceptions. When dealing with social 13

PAGE 14

issues such as perceptions and st igma, many factors could potentially influence an individual's perceptions, including their status as a parent, their socio economic status, their previous experiences with sex offenses and other crimes and exposure to the media. Furthermore, this project will examine the subtlety present in the comparison of "sex offenders" and "sex predators." A random assignment of those terms will be applied to the study to examine which terms is more salient to participants, and whether or not they are able to distin guish the difference in the two terms. All of these factors will be tested in conjunction with registry knowledge moral panic measures, legal support and accuracy to determine just how important and concerning the presence of sex offenders is in the eyes of the general citizenry. 14

PAGE 15

CHAPTER 1 INTRODUCTION Moral panics as a phenomenon exist because there is a social problem occurring, typically deviant in behavior and associated with one villanized group of people. These social problems disrupt our societal norms and cause alarm among our citizenry. Tradit ional moral panic literature makes a clear distinction between the prominent members of society and "others" in other words, there is a division between "us" and "them." Stanley Cohen's original work on moral panics identifies the "other" group as a "fo lk devil," a group to be alienated and sometimes feared (Cohen, 1972). This allows for the regular citizenry to point their finders and clearly identify the group that is to blame for the problem at hand. This finger pointing can last for years and has t he potential for serious legal ramifications for those receiving the blame. Race, religion and economic status have all caused the distinction between groups. Sometimes, those who fall into the "them" category are villanized because of their il legal actio ns. The idea of a moral panic sweeping the country is nothing new. Over the centuries, there have been multiple groups targeted as dangerous groups. In America alone there have been several moral panics as seen by the witch hunts in New England, Prohibi tion in 1920's America, fears of "reefer madness" associ ated with marijuana in the 1930 s, the red threat of communism in the 1950 s and the War on Drugs in the late 1970s and 1980 s (Goode & Ben Yehuda, 1994). Some might even argue that there is a current m oral panic surrounding the issue of obesity in the United States (Campos, Saguy, Ernsberger, Oliver and Gaesser, 2006). All of these panics were the result of one targeted group that was identified as a threat to mainstream culture. While moral panics ha ve extended to many things drugs, homosexuality and 15

PAGE 16

communism there has been a recent alarm sounded in regards to sex offenders and the threat that they pose to t he community as a whole, but es pecial ly in regard to children. This vilification transcen ds all social categories, including race, gender, ethnicity and region. Many individuals are outcast because of they engage in deviant actions or because they pose a threat to all non deviant citizens. This way of thinking introduces the idea of a moral panic into mainstream society. The panic surrounding sex offenders and the registry is not the first moral panic to take place surrounding the sexual activities of individuals. In the past there have been moral panics regarding homosexuals and "sexual de viants" (Goode & Ben Yehuda, 1994). During the 1930 1950 time period, the United States was worried about the sexual deviants and the sex fiends that posed a threat to families everywhere (Goode & Ben Yehuda, 1994). Local newspapers covered the sexual as sault cases and portrayed the perpetrators as sexually deviant individuals. In their book on moral panics, Goode and Ben Yehuda give the example of a murder of a young girl that took place in Los Angeles in 1949 (1994). The body is found mutilated and a city wide man hunt ensued to find Fred Stoble, the man who killed the little girl. The arresting officer was praised as a hero for being the "capturer of the sex fiend" (Goode & Beh Yehuda, 1994; 18). The story of Fred Stoble is only one of the highly co vered sexually related murders between the 1930 1950 time period. However, the murder took place during a time of heightened responsive ness to high profile murders. Drawing a parallel to the current state of moral panic surrounding the sex offender registry, the 1930 1950 moral panic over "sexual psychopaths" developed 16

PAGE 17

quickly, leaving a series of protective laws in its path (Goode & Ben Yehuda, 1994; 18). Goode an d Ben Yehuda explained that a pattern can be found in most states in regards to the passage of the "sexual psychopath" laws that begins with "a few serious sex crimes committed in quick succession" which are all high profile in nature (1994; 18). Once the se crimes are known to the masses, more victims of sex crimes were likely to come forward thus increasing the spread of the panic (Goode & Ben Yehuda, 1994). The bills are passed and later amended, but in some of the research concerning the sexual psychop ath laws, Sutherland concluded that neither the fear nor the passage of law is directly related to the increasing amounts of sexual crimes (Sutherland, 1950). One of the features of moral panic is the disproportionate response to the specific problem. Th is disproportionate response is due to the idea that public perceptions do not always match the findings of empirical research. Rather the fear and the laws seem to play off of one other, guiding the panic forward without being informed by research (Good e & Ben Yehuda, 1994; Sutherland, 1950) This example showing the fear associated with sexual psychopaths is a good starting point for the development of this paper. Although the media coverage in the 1930 1950 time was not as extensive as it is today, th e newspaper coverage of these cases played a large role in promoting the panic over sexually deviant individuals. Researchers state that media exploitation enables a panic to catch on quickly (Garland, 2008). It is the original work by Stanley Cohen (197 2) from which we get the development of a moral panic and how his work was necessary in describing what has taken place in cultures for many years. Cohen's work allows for the systematic identification of a moral panic all of the previously discussed ex amples fit within the 17

PAGE 18

mold developed in his book, Folk Devils and Moral Panics (1972). Before the discussion begins in connecting the moral panic literature to the sex offender registry, it is important to discuss the hallmarks of a moral panic Cohen's work allowed for the labeling of the participants of a panic and placed a clear emphasis on the idea that these panics are reoccurring events that take place throughout history; sometimes they last and sometimes they don't, but they are a very real phenomenon when they occur. In his seminal book, Cohen discusses that "societies appear to be subject, every now and then, to periods of moral panic. A condition, episode, person or group of persons emerges to become defined as a threat to societal value s and interests; its nature is presented in a stylized and stereotypical fashion by the mass media; the moral barricades are manned by editors, bishops, politicians and other right thinking people; socially accredited experts pronounce their diagnoses and solutions; ways of coping are evolved or (more often) resorted to; the condition then disappears, submerges or deteriorates and becomes more visible. Sometimes the panic passes over and is forgotten, except in folklore and collective memory: at other time s it has more serious and long lasting repercussions and might produce such changes in legal and social policy or even in the way society conceives itself" (Cohen, 2004:1). Cohen's definition calls attention to the issue of transiency some panics are fl eeting and some have long last ing legal and social repercussions. These r epercussions represent the back end of the panic. Similar to the perspective that Edward Sutherland took regarding sexual psychopaths and the laws of the 1930s this research springs from the sex offender registry and the laws surrounding the supervision of sex offenders. By examining the legal reaction to sex crimes, this paper aims to show that the moral panics literature is somewhat incomplete and needs to be reformulated surround ing the lasting legal effects of a panic. The majority of the literature focuses on the presence of the actual panic itself, but more needs to be written in regards to the post panic portion of Cohen's definition. Specifically, this study 18

PAGE 19

will examine th e "serious and long lasting repercussions" and the "legal and social policy" associated with sex offenders. The currently suggested moral panic extends to the supervision of sex offenders on the registry and the present laws that keep them supervised by th e state. There is great debate about whether or not the current laws are harsh enough or whether they are too strict Furthermore, there is a body of literature that suggests that sex offender registries are more of a feel good type of legal policy that does not really show much effectiveness, but is still supported because the general citizenry feels safer with these laws in place ( Adkins et al., 2000; Avrahamian, 1998; Gonnerman, 2007; Levenson and Cotter, 2005; Levenson, D'Amora and Hern, 2007; Madden, 2008; Minnesota Department of Corrections, 2003; Petrosino and Petrosino, 1999; Prescott and Rockoff, 2008; Sandler et al., 2007; Schram and Milloy, 1995; Veysey, Zgoba and Dalessandro, 2005; Welchans, 2005 ). People want to know what types of sanctions are enough to keep sex offenders away from our families and what level of supervision will keep our children safe from perceived chronic offenders (Levenson, Zgoba and Tewksbury, 2007 ) These questions are ones that politicians, law enforcement and even p arents ask themselves on a regular basis especially every time a new threat is posed to their community. When high profile sex crimes are committed, the fear is renewed among parents. This fear leads to the quick legislative response to crack down furt her on sex offenders. While the progression of sex offender registry laws will be discussed further in the paper, it is important to affirm their presence in relation to the moral panic sweeping the nation. 19

PAGE 20

What we see with sex offenders is that there is already a legal system in place and that laws have already been implemented to control and regulate this "folk devil" group. The legal and social changes references in Cohen's definition (2004: 1) play an important role in why the sex offender moral pan ic still persists. There is no way for this panic to die out because the registry keeps us aware of this so called threat. Social media, news organizations and the Internet have ignited an explosion of opportunities for individuals to be informed of sex offenses on an immediate basis. In addition, the registry is only just a mouse click away. If citizens are not actively pursuing the online version of the registry, some local branches of law enforcements have brought the registry into our homes via the Sunday newspaper. In September 2012, the Alachua County Sheriff's Office printed and distributed an annual booklet of all the sexual offenders and predators living in the county (ACSF, 2012). There is also a PDF version available online through the Gain esville Sun website (Gainesville Sun, 2012). This publication once again reinforces the idea that citizens should be aware of the threatening group of offenders that live amongst us, thus not letting the panic die away. This paper will examine perceptions of the sex offender registry and those registered on it, from a post panic perspective. Furthermore, t his paper will make an argument for why the elements of a true moral panic still persist in regards to sex offenders. Additionally, the post panic lega l changes have caused some unintended consequences such as difficulties finding housing and steady employment. Furthermore, the registry laws will also be reviewed to show the relationship between high profile sex crimes and the legislation that has been passed to increase the severity of sex offender laws. Prior literature will be reviewed for all of these areas. Finally, the 20

PAGE 21

current project is proposed that examines community members throughout the country and their perceptions of sex offenders and th e sex offender registries in their state 21

PAGE 22

CHAPTER 2 LITERATURE REVIEW Moral Panics Literature In his development of the moral panic ideas, Cohen discusses the need for a vill ain that needs to be feared a group of persons emerges to become defin ed as a threat to societal values and interests" (Cohen, 2004: 1). Cohen calls this villainous group the "folk devil" (Cohen, 1972; Goode & Ben Yehuda, 1994). Because the villanized group exists, communities are then able to exaggerate the negative infor mation associated with the folk devil. The folk devil is used as a scapegoat for communities, allowing individuals to use the deviant group as an example of what behaviors are wrong and go against the societal norms, essentially labeling the group as "the m;" people not to be trusted and individuals who threaten the very heart of our culture (Goode & Ben Yehuda, 1994). Whether or not the villanized group actually poses a threat to society, as a whole is not the issue. If society believes that a threat is present, then the moral panic has the ability to continue and thrive (Cohen, 1972; Goode & Ben Yehuda, 1994). Cohen compares the moral panic of an issue to an epidemic or an illness (1972). The panic starts to permeate society and quickly the idea is fo rmed that "something should be done" about the folk devil's behavior (Goode & Ben Yehuda, 1994: 31). But before this illness can be formally called a moral panic, a set of elements must exist according to Cohen. He lists the five elements as follows: con cern, hostility, consensus, disproportionality and volatility. The first element that Cohen proposes is elevated concern over an issue (1972, Goode & Ben Yehuda, 1994). In other words, how aware is the community that a threat may exist? If community mem bers are aware of the 22

PAGE 23

possible threat, how worried are they that the threat may affect their own lives? The best way to measure how concerned communities are about a subject is to ask them directly (Cohen, 1972; Goode & Ben Yehuda, 1994). Public opinion surveys and polls are a way to get the truth about how concerned people really are about the issue at hand. However Best makes the point that concern is n ot the same thing as fear, but rather f ear takes place when people think that a threat is imminent a nd has a good chance of occurring (1990) Its focus is on victimization. Concern is similar to fear but is thought of as a more preliminary version of the emotion there is a chance that something may happen but right now the situation only needs to be monitored (Best, 1990; Goode & Ben Yehuda, 1994) 1 Cohen's second essential element of moral panics is the level of hostility that the majority o f community members feel toward the threatening group. It is within this condition that members of the folk devil group are labeled as the enemy to society (Barlow, 1993; Cohen, 1972; Goode & Ben Yehuda, 1994). Hostility felt by community members extends to the threatening behavior, b ut also a certain group is usually attributed responsibility for the behavior (Goode & Ben Yehuda, 1994). The hostility felt then allows the members of society to separate themselves from the folk devil group in other words makes the situation one of "u s and them" (Cohen, 1972). Most often, the deviant group is singled out on the basis of social categories or criminal activity (Barlow, 1993). For example folk devils were made out of marijuana users in the 1930 s and of crack cocaine users in the 1980 s. The behavior of drug use was villanized, but the drug 1 Ferraro and LaGrange (1987) suggest that fear of crime is "an emotional response of dread or anxiety to crime or symbols that a person associates with crime." This emotional response is also present in regard to sex crimes, that can translate into increa sed legislative advancements that levy punitive restrictions on registered sex offenders. 23

PAGE 24

users were made out as people to be feared. The vilification can only feed a moral panic when there is societal agreement; the moral panics would not be able to grow without social support Societal consensus was developed as the third element necessary in the explanation of Cohen's moral panic argument. The consensus argumen t illustrates the need for over arching agreement among communities (Goode & Ben Yehuda, 1994). Just how much consensus is nece ssary for a moral panic to take place? Research has shown that the numbers are not able to say when exactly a moral panic has been realized (Cohen, 1972; Hall et al., 1978; Zatz, 1987); however, there must be majority agreement among the group being measu red (Goode & Ben Yehuda, 1994). In other words, if you are measuring consensus among a neighborhood, mass agreement must be similar to the mass agreement felt by an entire state or even the nation. Using 1920's Prohibition in the Unites States as an exam ple, it could be argued that there was a large constituency of people who felt that alcohol was a threat and that there was a large constituency of people who felt that alcohol was not a threat. However, those who felt it was a threat were more organized in their concerns, were in more agreement that steps should be taken to do something about the threat of alcohol and were more influential over policy makers at the time (Goode & Ben Yehuda, 1994; Gusfield, 1955; Sinclair, 1962; Kobler, 1973). In order fo r consensus to be strengthen the proponents of Prohibition needed to create a crisis surrounding the threat. Therefore, the proponents created examples full of depraved men who drank to excess, and beat their wives. If women continued to drink they would become sexually perverted (Kobler, 1973). Minorities would become unmanageable and without the action of politicians, alcohol 24

PAGE 25

would destroy our country (Goode & Ben Yehuda, 1994; Kobler, 1973). Using these examples as a way to scare citizens, it was eas y to pass a constitutional amendment outlawing alcohol in the United States. So by that example it can be seen that even though there may be a consensus felt by some individuals it is necessary to increase that consensus by stirring the waters and incre asing fear among the general population. Now that communities are afraid of the threat, they will come together in unison against it. That consensus paves the way for action to be taken, which furthers to strengthen the moral panic However, sometimes t aking action can further make the argument that the threat is being overblown and that people are disproportionately over reacting. This could be seen when Prohibition was overturned only a few years after it was implemented. Cohen's fourth element needed in the construction of a moral panic is the disproportionality of the societal reaction to a threatening behavior or group. Davis and Stasz explain disproportionality as any action that is taken in response to the threatening behavior that is "above and b eyond that which a realistic appraisal could sustain" (1990: 129). T he social reaction is seen as an extreme overreaction to the threatening behavior, giving the threatening behavior more credit than it may deserve (Goode & Ben Yehuda, 1994; Jones, Gallag her and McFalls, 1989). One way to assess whether a reaction is disproportionate or goes beyond what a realistic appraisal could sustain, is to compare it with the empirical evidence available about the problem. T he social reactions may create more harm than good in trying to rectify the threatening behavior (Goode & Ben Yehuda, 1994). These actions may have a negative side effect to them; however, when enacted they were thought to be the 25

PAGE 26

accurate solution to the problem (Goode & Ben Yehuda, 1994; Jones Gallagher and McFalls, 1989). The confidence associated with the proposed solution allows for the disproportionality of the reaction to be negated (Cole, 1993; Goode & Ben Yehuda, 1994). The issue of a disproportionate response to the threatening grou p and behavior te nds to occur because of a quick formed response; most often the public calls for a retaliation response without allowing for time to examine the side effects associated with the implemented actions (Goode & Ben Yehuda, 1994). This volatil e response helps explain why sometimes the response actions may not al ways be the correct responses. Therefore, Cohen advanced a final element of moral panic. It is the volatility of the community feelings over the threatening behavior or group (1972). H ow quickly does the heated arguments and feeling burn out over the concerning situation? As previous literature has stated, moral panics are often a quick forming reaction to a behavior that takes place and is conducted against a perceived threatening gr oup (Cohen, 1972). As a result of the moral panic, typically some retaliatory behavior is conducted on behalf of the community against the folk devil, as stated earlier. The retaliatory behavior of the community is also quick to be implemented in terms of legislation and enforcement of the law based on the reaction of the social majority (Goode & Ben Yehuda, 1994). Using the example of the 1970s/ 1980s war on drugs, it can be seen that drug use was always publically known. But, many of the attempts to c urb drug use in the past were not very effective in reducing drug use, sale or distribution. At the end of his term in office, President Nixon began some crackdowns on the drug industry in the United States, but also sparked the newest panic over drug use (Goldberg, 1980; 26

PAGE 27

Goode & Ben Yehuda, 1994). In the 1980 s the drug of choice was cocaine and a very identifiable group was using more crack cocaine than other sections of society (Goldberg, 1980). Eventually, the war on drug s shifted focus to other drugs and eventually lost popularity by the mid 1990 s, providing that the volatility of the war was simply a flash in the pan as far as moral panics go. These five elements combined form Cohen's definition of a moral panic. In order for a moral panic to be applied to a specific situation, the five elements must be able to be explained and shown to fit the behavior of community members. While this project explores to see if a moral panic is taking place, the main focus of this paper centers on the post panic activity. It is hypothesized that all five of Cohen's elements are present regarding sex offenders and crimes in our society today, but in addition t here is a vast amount of legal change that has taken place regarding the regulation supervision, and control of sex offenders. These legal changes have been implemented and increased in severity every time a high profile celebrated child victimization case occurs. Of the national laws that regulate sex offender registries and supervision, three of the biggest ones are named after infamous cases Jacob Wetterling ( Public Law 103 322 ), Megan Kanka ( Public Law 104 145 ) and Adam Walsh ( Public Law 109 248 ) These "memorial criminal justice policies" (Surette, 2010: 2). There have been additional state laws also named after victimized children. These pieces of legislation have arguably made life difficult for sex offenders to reintegrate back into society In their attempts to supervise sex offenders, the laws have also made the general public afraid and confused. The Adam Walsh Act ( Public Law 10 9 248) in particular has added significantly to the legal restrictions placed 27

PAGE 28

on sex offenders. This act made the nationwide registry system more cohesive but also left open a lot of interpretation open for states as long as states were in compliance with the federal laws, they are allowed to place additional requirements and restrictions on their sex offenders (Public Law 109 248). Plus, there is no standard classification system for registered sex offenders and this can cause confusion to the general public, which does not know the legal stipulations imbedded in the laws. T hese laws have also made it difficu lt for regular community members to ignore the registry and the presence of sex offenders. Due to the efforts of law enforcement, the news media, social media and politicians, society is not allowed to forget about the potential threat that sex offenders pose. The laws continue to punish these offenders as a deterrent factor but the recidivism rate has already been historically low (Finkelhor and Jones, 2012; Lonsway and Archambault, 2012; Mancini, 2013) even bef ore these laws went into effect Therefo re, they may only serve a feel good purpose for regular community members (Adkins et al., 2000; Avrahamian, 1998; Gonnerman, 2007; Levenson and Cotter, 2005; Levenson, D'Amora and Hern, 2007; Madden, 2008; Minnesota Department of Corrections, 2003; Petrosi no and Petrosino, 1999; Prescott and Rockoff, 2008; Sandler et al., 2007; Schram and Milloy, 1995; Veysey, Zgoba and Dalessandro, 2005; Welchans, 2005 ). It has been argued that punishment serves an expiation purpose an offender must be punished in order for his/her wrong to be ( Durkheim, 1884 /2011 ). Durkheim makes the point that many criminal behaviors are also immoral, but he questions 28

PAGE 29

whether or not it is really our job as a society to improve the guil ty person. After all he argues the guilty ma n cannot be improved through punishment (Durkheim, 1884 /2011 ). It is the role of our government to lead and protect for as long as that government is properly performing its duty (Jones, 1999 ). The following review of the literature surrounding sex offe nders and the registry questions whether or not the government is not doing an adequate job of maintaining a safe and balanced form of supervision and control regarding sex offenders. These laws severely limit the amount of opportunities that sex offender s have in reintegrating back into society. Employment and housing are tw o of the biggest issues that are restr icted by the registry laws (Jenkins, 2001; Levenson and Cotter, 2005; Levenson, Zgoba and Tewksbury, 2007; Ost, 2002; Tewksbury, 2004, 2005; Tew ksbury and Zgoba, 2010; Zanbergen and Hart, 2006; Zevitz and Farkas, 2000 ). Based on what is seen in the law and in the academic literature, this paper will examine the realities associated with the registry and the departure the current system has taken from a well intent ioned supervision structure to a version that may be ineffective and essentially only feel good in nature. Connecting Moral Panics and the Sex Offender Registry The litera ture on moral panics establishes that one of the main premises be hi nd the panic is the mass over exaggeration by the community in regards to the actual events that took place (Cohen, 1972; Goode & Ben Yehuda, 1994). The exaggeration leads to distortion for example if one child was abducted and fondled by an adult a mo ral panic would distort the event in a violent kidnapping and rape committed probably by a multiply convicted sex offender through various channels of communication. 29

PAGE 30

Cohen argues that the general public must believe that the distortions are actual truths in order for the panic to occur (1972). As explained earlier, Cohen's development of the moral panics argument states that there are five necessary elements that are needed for a moral panic to take place concern, hostility, consensus, disproportional ity and volatility (1972). The five elements can be linked to the sex offender registry to show that a moral panic concerning sex offenders currently exists. Sex offenders are cast as the "folk devil" and are associated with potentially harmful future be haviors. T he community perceives that they have committed sex crimes in the past and have the potential to do so again (Angelides, 2003; Levenson, Zgoba and Tewksbury, 2007, Ost, 2002). As discussed in detail later, the empirical evidence does not suppor t this characterization of most sex offenders. It is possible t hat their thinking is distorted? Due to this distortion, when community members hear the term "sexual offender" their mind s might jump to "sexual predator" instead. This kind of distortion fits well with Cohen's discussion of the folk devil Seeing all sex offenders as t hreatening and perceiving the registry laws as a basis for their control contributes to how a moral panic emerges and is sustained. Cohen's first element concern speak s to the worry among community members in regard to potentially threatening behaviors that is associated with a group. In connecting concern to the sex offender registry, we see that community members maybe especially parents have concern over what se x offenders are doing within neighborhoods (Angelides, 2003). Again, there may be a gap between perceptions and empirical evidence While the sex offender registry provides for community protection and is used as a tracking agent, sex offenders still are capable of committing additional 30

PAGE 31

sex crimes (Levenson, Zgoba and Tewksbury, 2007, Ost, 2002). For parents, the registry symbolizes an important form of protection for their children, despite the criticisms that are associated with the registry itself (Le venson, Zgoba and Tewksbury, 2007). But even with the current state of the registry and the highly punitive restrictions that are placed on sex offenders today, parents are still worried and concerned for their children's safety. The second element of a moral panic is hostility that is felt by the societal majority against the threatening group. Contributing to the hostile feelings that community members feel towards sex offenders, the media exaggerate the immediate threat that sex offenders pose to fam ilies (Levenson, Zgoba and Tewksbury, 2007). Whenever a sex crime occurs, the media sometimes embellish th e facts of the story and report the worst information that they have about the case. These embellishments help fuel the fear that people have about the threat posed by sex offenders, despite the fact that some sex crime rates have decreased since the 1990s (Finkelhor & Jones, 2004; Levenson, Zgoba and Tewksbury, 2007; Maguire & Pastore, 2003). The hostility must be felt broadly across community memb ers in order for the moral panic to take place. This suggests the important of Cohen's consensus element and that it must be present. So for the purposes of the moral panic connection to sex offender registries, there must be consensus Parents of the v ulnerable children, specific neighborhoods and previous sex crime victims all fall within the scope of individuals who must be in agreement over the posed threat (Jenkins, 2001; Levenson, Zgoba and Tewksbury, 2007; Ost, 2002). So even if some do not feel as though sex offenders pose much of a threat to their safety, the vast majority of the community will 31

PAGE 32

perceive a serious risk involved with convicted sex offenders living so close by. Previous research would suggest in this instance, a moral panic is tak ing place in the concerned community (Goode & Ben Yehuda, 1994; Hall et al., 1978; Zatz, 1987). The next element needed for the moral panic to be present is the disproportionality of the societal reaction to the perceived threat (Cohen, 1972). As noted be fore, disproportionality speaks to the notion that the societal reaction goes above and beyond what is what a reasonable appraisal would allow This overreaction is mismatched with what the empirical evidence shows In the instance of the sex offender re gistry, the disproportionality may speak to the extreme restrictions that are placed on the registrants. Residency restrictions, job restrictions and supervision of the offender all serve as tools to keep sex offenders on a short leash (Jenkins, 2001; Lev enson, Zgoba and Tewksbury, 2007; Ost, 2002; Tewksbury, 2004). Residency restrictions in particular make it difficult for sex offenders to maintain housing since they cannot live within certain exclusionary zones such as areas close to schools, parks an d bus stops (Levenson, Zgoba and Tewksbury, 2007; Tewksbury, 2004). Making the link back to the distorted thinking patterns of community members, this disproportionality might be in response to sexual predators not the offenders. The media dramatizes and highlights the heinous crimes that occur, making it seem as though the majority of sex offenders are predatory in nature. The reality is that sexual offenders vastly outnumber sexual predators in most states. For exampl e in Florida, sex offenders out num ber sexual offenders with a ratio of over 5 to 1 2 (FDLE, 2012b ) 2 As of 10/24/2012, there were 49,776 sexual offenders and 9,725 sexual predators living in Florida according to the Florida Department of Law Enforcement. That prov ides a ratio of 5.11 sexual offenders for every 1 sexual predator (FDLE, 2012b). 32

PAGE 33

Due to the nature of the media coverage and from the knowledge that people believe is correct, it is not an outlandish assumption that many citizens would not believe this statement about the numbers of predators, because it is not what we have grown to expect. In terms of disproportionality, the laws and restrictions are really being put in place for the predators not the offenders but the nature of the catchall laws lump everyone together in the minds of community members Whether or not the current registry laws are the laws needed to keep communities safe from sex offenders, the extreme res trictions placed on sex offenders could be construed as being an overreaction on the part of scared, nervous community members especially given the low rate of recidivism The final element needed to complete the moral panic hypothesis is the notion of volatility. Vola tility is measured as the quick rising reaction to a threatening act that takes place within a community (Goode & Ben Yehuda, 1994). Also involved with the volatili ty measurement is the fast burn out associated with the feelings of being t hreatened (Cohen, 1972). In other words, are people quick to react with anger about a situation and after a while, cease to be concerned? Relating the element of volatility to the sex offender re gistry, one could see the quick rising reaction of the com munity after a high profile child abduction, sexual assault and subsequent murder takes place (Caldwell, 2007; Harris, 2006; Sample & Bray, 2003). These high profile cases are representative of a more predator style sex offender rather then the lower grad e type of offenders, which are more common (FDLE, 2012 a ). Media coverage exacerbates these quick responses to predatory crimes and together we see a high grade response to a potential ly low grade problem. The wide net lumps a lot of individuals together, many of 33

PAGE 34

who do not commit the heinous crimes demonstrated in later sections of this paper. With these abductions, a new wave of anger takes hold of communities often resulting in increased sex offender registry legislation (Baron Evans & Noonan, 2006; Ca lley, 2008; NCSL 2008). Afterwards, the quick rising anger dies down and normality returns to the community, with individuals believing that the new legislation is keeping them safe. Then another high profile abduction takes place and the cycle begins o nce again. The dissertation examines the existence of a connection between moral panics and the sex offender registry. Predominately, the study will focus on the panic itself and the legal ramifications that exist post panic, which sustain the potential for a panic mentality Due to the nature of the sex offender laws, there is no opportunity for the sex offender panic to t ruly die down and fade into the annals of history. The immediacy of the threat might be diminished any time there is a new case tha t occurs, the panic reasserts itself and is reinvigorated. This can also be seen by the way citizens are chronically reminded of the potential threat that sex offenders pose news outlets, social media and law enforcement make sure that this panic is ali ve and continuing When the media discuss these types of offenses, they tend to dramatize the situation as posing a threat (Hollinger and Lanza Kaduce, 1988). No m atter what the offense, there may be a tendency to classify sexual offenders as sexual pred ators, wh ich can be a misclassification and distortion of the situation. The next section of the paper describes the high profile cases that served as the foundation for many of the sex offender laws. These cases were admittedly brutal and heinous, but have become almost celebrated keystones for why we should be afraid of sexual offenders. In fact they epitomize the celebrated case ( Walker, 2001 ). T hese 34

PAGE 35

media reports contribute to the general public afraid of sexual p redators. There is such a wide net cast over these individuals, grouping them together This distortion of labels, t he media's response to these cases and the speed of the legislative bodies have created a harsh and severe post panic legal arena where sex offenders have a difficult time t r ying to re enter society. The media in particular significantly advances this fear by pandering to the side of us that feels most threatened (Hollinger and Lanza Kaduce, 1988) By playing upon our vulnerabilities and incorporating a little bit of exagge ration, news corporations can sell stories ( see Su rette, 2011 ) The drama that is added to news reports, allows "the media [to] contribute to public perceptions of a threat, regardless of accuracy," (Hollinger and Lanza Kaduce, 1988: 114). For example, t here have been local news reports about Internet sex stings and other arrests that have been made for sex offenses. However, no matter what the offense may be, the media group all of these defendants under the "predator umbrella, meaning that all sex off enders and even some newly arrested individuals with no previous criminal records are grouped in this category (Silman, 2012; WCJB 20, 2012) This labeling precedes the actual case processing they have not been convicted and are not on the registry, so they cannot be identified as a sexual predator. However, the label makes the news more exciting which in turn continues to promote the threat and the panic over sex offenders. High Profile Abduction Cases Lead to New Laws Beginning in the early 1990s, a string of high profile child abductions took place causing a nation of parents to fear for their children's safety. While some of the children were found, some were sexually abused and later killed. The deaths of these chi ldren further enraged the nation and resulted in the passage of federal and state legislation. 35

PAGE 36

This legislation created a national sex offender registry and mandated that the states follow suit. The first high profile case that resulted in legislation wa s the abduction of Jacob Wetterling, an 11 year boy from Minnesota who was abducted on his way home from school by a masked gunman. Jacob and the gunman were never heard from again. His body was never found but he is thought to be dead as a result of the abduction (Jacob Wetterling Resource Center, 2010). Jacob's abduction resulted in the first piece of sex offender registration legislation to be passed at the state level (Michigan State Police, 2006). According to the law, a person who is convicted o f a criminal offense who is a minor, who is convicted of a sexually violent offense, or who is a sexually violent predator" must register with the state in which they reside (Jacob Wetterling Act, 1994). The biggest, and possibly best known, case to affe ct the sex offender registry on a national level was the abduction and eventual murder of Megan Kanka in New Jersey (Megan Kanka Foundation, 2001). Megan was lured into the home of her neighbor in 1994, with promises that she would get to see the neighbor 's puppy. That neighbor, Jesse Timmendequas was a convicted sex offender who raped and murdered Megan once she was in his home (Megan Kanka Foundation, 2001). Within ninety days of Megan's murder, the citizens of New Jersey rallied together, signed a pet ition for a registration law to be passed. New Jersey legislators signed the bill for Megan's Law into effect, ensuring that every family will have the "right to know that if a dangerous sexual predator moves into their neighborhood" (Megan Kanka Foundati on, 2001). Megan's Law, as it most commonly known now, laid the foundation for the national sex offender registry (Michigan State Police, 2006). Essentially amending the Jacob 36

PAGE 37

Wetterling Act, Megan's Law required "all 50 states to establish and maintain a community notification system" (Shiavone and Jeglic, 2009: 680; Center for Sex Offender Management, 2001). After the murder of Megan Kanka, several other children were abducted, sexually assaulted and subsequently murdered. While many individuals have e xperienced the same fate as the most well know n cases, only a few laws have been attached to the murdered children. Pam Lychner was the next high profile case that received attention but was an exception to the child abductions. At 31 years old, Pam beca me the victim of an attempted rape by a workman who was at her house to fix a set of leaky pipes (Bureau of Justice Assistance, 2010). Before the man could complete the rape, Pam's husband Joe came home and stopped the man from completing the act. Police found out that man was convicted rapist and child molester. The Pam Lychner Sexual Offender Tracking and Identification Act was passed in 1996, shortly after Megan's Law, which established more requirements for the registries including lifetime registrat ion, for severe and repeat sex offenders (Bureau of Justice Assistance, 2010; Michigan State Police, 2006). In 2000, The Jeanne Clery Disclose of Campus Security Police and Campus Crime Act was passed to modify the existing regulations established in Meg an's Law. The Clery Act was passed after the rape and murder of 19 year old Lehigh University freshman Jeanne Clery (Lehigh University, 2006). The Clery Act made it mandatory for college campus to alert students and families about registered sex offender s residing on campus and attending classes, as well as publish the statistics gathered about crimes happening on college campuses (Lehigh University, 2006; Michigan State Police, 2006). 37

PAGE 38

All college campuses are required to comply with the Clery Act and st atistics can be found on the university's police website (University of Florida Police Department, 2010). In 2005, Florida Governor Jeb Bush signed the Jessica Lunsford Act into law. The act was created after Jessica was abducted from her home in Homosass a Springs, Florida. She was raped and later murdered by John Couey, a man who worked at her school as a service contractor and who was also a convicted sex offender (Florida Department of Education, 2005). The new act requires that schools conduct more t horough background checks on individuals who are working for public or private schools to ensure the safety of the children. These background checks will help to keep potentially dangerous individuals off of the school grounds and away from the children ( Florida Department of Education, 2005). In 2005, the Florida law was proposed at the federal level, but was never passed by Congress (Florida Department of Education, 2005). The most recent federal legislation to be passed is Adam Walsh Child Protection and Safety Act of 2006, named for the son of television personality John Walsh. John's son Adam was abducted from a Sears store in Hollywood, Florida in 1981. He was abducted by Otis Toole, a convicted serial killer, who later confessed on this deathbed that he was the murderer of Adam. On August 10, 1981 two fishermen in Vero Beach, Florida found Adam's severed head (MSNBC News Services, 2008). After Adam's death, John Walsh began his work with the National Center for Missing and Exploited Children. The passage of the Adam Walsh Child Protection and Safety Act of 2006 provided an opportunity to create a national, cohesive and more systematic form of sex offender registration (Calley, 2008). Under this act, all states are required to create a 38

PAGE 39

sex offe nder registry that is compatible to the national database registry (NCSL, 2008). The creation of this law in effect eliminated the statute of limitations for most sex offenses (Baron Evans & Noonan, 2006). Therefore, convicted sex offenders who did not h ave to register in the past were now required to register, even though they never had to before (Baron Evans & Noonan, 2006; NCSL, 2008). The act states that states must comply with the registry rules and must do so by 2009 or lose state funding as a resu lt (Adam Walsh Act, 2006). As of July 2011, only a handful of states and Indian Tribes have complied with the act's requirements. These states and tribes include Delaware, Florida, Michigan, Nevada, Ohio, South Dakota, Wyoming, Guam, the Confederated Tri bes of the Umatilla Indian Reservation, the Confederated Tribes and Bands of the Yakama Nation and the Grand Traverse Band of Ottawa and Chippewa Indians in Michigan (Department of Justice, 2011). The reason for discussing the individual laws is to show ho w after a high profile sex crime and subsequent murder takes place, the public is quick to demand new legislation to restrict the activities of sex offenders even further. In essence what these laws are trying to restrict is the activities of sexual pred ators T his section of the paper shows a quick legislation change is representative of the volatility element of the moral panic hypothesis where even with sustained concern about sex crimes there are ebbs and flows in public sentiments Again these law s are targeted to protect us from the sex ual predator, not the sex ual offender. The blurring between this distortion causes confusion among community members. Despite what the literature may tell us, parents view t hese restrictions as necessary (Potter & Potter, 2001). If the laws are severely restricting sex offenders, 39

PAGE 40

then they are not able to abduct any more children. It is argued that even if children try to keep themselves safe and do not actively interact with strangers, the sex offenders will fi nd the children (Potter & Potter, 2001). Regardless of the legitimacy of the arguments made by Potter & Potter (2001), the fear held by parents is legitimate causing the adaptations to sex offender registry laws to be passed with little opposition. As stated earlier, the quickly passed laws are representative of the volatility element of a moral panic. The kno wledge of registration laws has implications for moral panic analysis. If knowledge is low, there is room for volatility in reactions when these tragic cases are publicized. With high levels of knowledge, a less emotional response would be expected, which also holds implications for t he disproportionate reaction represented by the ever increasing severity of legislation. Those with higher levels of knowledge would appreciate the diminished returns for protection that yet another registry law has. Community Perceptions of Sexual Offe nders Sex offenders have often been vilified as folk devils that citizens should be fearful of. From t he prior review of the sex offender registry laws, it can be seen that there is a highly emotional response t hat takes place regarding sex offenses. Typ ically, parents fear sex offenders because they are concerned that their children may be victimized next. This emotional response can become overwhelming and cause citizens to take a punitive stance towards sex offenders. M any citizens are in favor of ha rsh punishments and punitive actions to be taken against known sex offenders (Roberts, Stalans, Indermaur and Hough, 2003). However, only a small amount of research has focused on just how supportive community members ar e for specific offender types (e.g. child pornography users or prostitutes or rapists) rather than for sex offenders as a whole 40

PAGE 41

(Roberts et al., 2003; Mears, Mancini, Gertz and Bratton, 2008). Due to the perceived threat, the public and legislators alike are willing to endorse strict punis hment because of the way these crimes are stereotyped. In particular, when a sex offender has a child victim, incarceration length is longer than that of other types of sex offenders (Lam, Mitchell and Seto, 2010). In addition to the harsher sentences im posed on sex offenders, there has been evidence of community demand for increased use of the sex offender registry and for public notification of sex offenders living in neighborhoods (Bates and Metcalf, 2007; Lam, Mitchell and Seto, 2010; Mears et al., 20 08; Seto and Eke, 2005). Essentially, community members are calling for more regulation of sex offenders by way of the criminal justice system (incarceration length) and by the state (monitoring through the sex offender registries once released back into the community). The research conducted on comm unity member perceptions of sex offenders supports the notion that there is a desire for harsher sanctions for these offenders who are considered to commit some of the most heinous crimes possible (Lam, Mitch ell and Seto, 2010). This support for harsher sanctions may stem from the belief s that sex offenders cannot control themselves and that their recidivism an almost guaranteed event despite efforts to rehabilitate them. Many citizens simply believe that treatment does not work for sex offenders (Sundt, Cullen, Applegate and Turner, 1998). Prior Research Conducted on Community Perceptions of Sexual Offenders Research has been conducted on community perceptions of child pornography users, which looks specifically at the age and gender of the child and the demographics of the child porn users to see if these variables affected community perceptions of the offender (Lam, Mitchell and Seto, 2010). Prior t o the Lam et al. study, the literature had 41

PAGE 42

shown t hat community perceptions were affected by the age of the child the younger the child, the more abusive the incident is perceived to be (Finkelhor and Redfield, 1984; Waterman and Foss Goodman, 1984; Maynard and Wiederman, 1997; Back and Lips, 1998). Th e issue of gender has had mixed results. Same sex incidents tend to be perceived as more abusive in comparison to opposite sex incidents (Maynard and Wiederman, 1997). In addition, adult female offenders and male child victims are perceived to be the lea st abusive of the different interactions (Broussard, Wagner and Kazelskis, 1991). In order to examine some the different variables that affect perceptions, Lam et al. designed a two part vignette study, which used fictional child pornography usage scenar ios (2010). The first vignette study varied the age and gender of the fictional minor to determine if this variation would "affect perceptions of crime severity, risk, and the probability that the offender was a pedophile" (Lam, Mitchell and Seto, 2010: 1 78). The researchers predicted that the younger the child, th e more severe the perceptions against the offenders would be. The results of the first vignette study showed that there was a significant effect present for the fictional child's age but not fo r the gender for the variable of perceived crime severity. The participants as a whole did not agree that the child porn user would reoffend when gender and age were varied. In addition, the age and gender variation for the child did not produce signific ant results that the offender was perceived to be a pedophile (Lam, Mitchel and Seto, 2010). The results of the first study are consistent with the literature that states age would affect the perceived severity of the offense. The issue of gender has pro duced mixed results in the past and did not produce statistically significant results in the first part of this study. 42

PAGE 43

The second vignette study varied the age and gender of the child pornography user to predict crime severity, risk and probability of ide ntifying the offender as a pedophile. Consistent again with prior literature, the researcher predicted that older users and male users would be perceived more severely. Lam et al. state that when there is a large age gap between the offender and the chil d depicted in the porn, perceptions tend to be more negative towards that offender. Female child porn users tend to be a rarity; therefore the negative perceptions are again linked to the male users in terms of severity of the offense, likelihood of recid ivating and identification as a pedophile (2010). The results of the second study did not find a significant effect for the offender's gender or age in terms of seriousness of the offense. However, male offenders were predicted to more likely to recidiva te compared to female offenders a finding that is consistent with the prediction made by Lam et al. There was no statistically significant difference between male and female offenders in terms of being identified as a pedophile. This finding did not fi t with the researchers predictions. When there is a specific stimulus presented, people may be able to differentiate among them and characterize situations differently. Those distinctions however are unlikely to be present when people generically think ab out the typical sex offense, as shown in the second portion of the Lam et al. study. Overall due to the results of the two studies, Lam et al. conclude that even through there is some separation between perceptions and empirical evidence, the results of t his first study may still hold significant weight in terms of public policy and the law. Since public opinions often times have an influence on the decision making of policy makers, there is a discrepancy on the treatment of sex offenders compared to non sexual offender (2010). Sex 43

PAGE 44

offenders are often subjected to harsher punishments, longer incarceration rates and must comply with registration requirements post release. However, because of the extremely negative public perception towards sex offenders, these increasingly tough punishments are not likely to change in the near future. The moral panic literature suggests the generic folk devil as the impetus for public concern, hostility, and punitive reactions that constitute a consensus. Other research h as examined community perceptions of sex offenses and the "get tough" measures taken to prosecute child pornography users (Mears, Mancini, Gertz and Bratton, 2008). It is suggested that those who offend against children are treated more harshly by citizen s and by the criminal justice system due to the long standing American ideal that children are innocent and must be protected, thus keeping the idea of "childhood" i n tact (Mears, Hay, Gertz & Mancini, 2007; Mears et al., 2008). Due to this notion, sex offe nses against a child are considered some of the worst offenses possible, but just how punitive should the punishments for chi ld molesters be? The most well known punitive actions taken against sex offenders include the statewide registries (Logan, 2003, T ewksbury, 2005; Mears et al., 2008) and civil commitment of sex offenders post release from incarceration (Sims & Reynolds, 2007; Mears et al., 2008). V oluntary chemical castration is infrequently used as a treatment method for rehabilitating sex offender s (Mears et al., 2008). Are these "get tough" efforts enough to keep sex offenders from further offending? In a nationwide telephone based survey, Mears et al., surveyed a random sample of 425 American adults, which they determined was representative of the U.S. population of adults when compared against census information (2008). The telephone 44

PAGE 45

survey asked respondents about a variety of sex offense related issues. The first asked about their opinion about sex crimes. In the aggregate, is there enough concern about sex crimes to make them a top priority for the criminal justice system? Follow up questions asked for opinions about the publicly available registry, residency restrictions and about incarceration of sex offenders who have adult victims. O ne other area of the study focused on whether or not participants would be in favor of using tax dollars to increase the amount of treatment that sex offenders received (Mears et al., 2008). Results from the telephone survey showed the majority of particip ants (54%) did feel as though the criminal justice system should make sex offenses a top priority, with increased support for policy efforts that focus on sex offend e r s. As for the three main follow up questions, the majority of participants supported use of the registries (92%), residency restrictions (76%) and incarcerating sex offenders (use generically to cover all sex related crimes) who commit their offenses against an adult victim (94%). These results shows that the most common forms of punitive se x offender policies really do coincide with public opinion (Mears et al., 2008). While these results do enforce the preconceived notions about sex offenders, the main focus of the study examines the proper way to punish individuals who commit sex offenses against minors. When given the option of sentencing the individual to probation, treatment or to pay a fine, an overwhelming 97% of participants decided that a prison or jail sentence was a more appropriate sentence (Mears et al., 2008). According to t he previous literature on sex offender treatment, most laypersons do not feel as though treatment is effective in the reform of sex offenders (Sundt, Cullen, Applegate and Turner, 1998). With that thought in mind, the Mears et al. study 45

PAGE 46

examined how likel y it would be for participants to support the idea of collecting additional tax dollars to pay for sex offender treatment (2008). The finding for that proposition showed that nearly half of the participants (48.2%) were unwilling to pay more in taxes to p ay for sex offender treatment even when the increase was as low as $25 a year. Based on the number of individuals who filed a tax return in 2005, the researchers estimated that if this tax increase were to occur, there would be more an $3.6 billion in t ax revenue that could go to sex offender treatment (Internal Revenue Service, 2006; Mears et al., 2008). This amount of money generated would substantially be able to help out the treatment programs across the United States in their rehabilitation efforts However, it would not be too far fetched for tax payers to argue that this amount of money would be better served helping out industry or provide opportunities for new jobs and that it should not go towards helping a group of individuals that most citi zens do not believe can be redeemed nor warrant redemption As the literature states, there are many community members who have very strong opinions regarding sex offenders who are capable of committing various types of sex crimes against both adults an d children. While o ur society feels highly punitive toward sex offenders, there is no other policy program that has received more notoriety than the sex offender registries in the respective states and at the federal level Community members also have st rong opinions regarding the registry, the level of punitive action taken against registrants, and whether or not the registry is truly guarding and protecting community members and their families. In this review on the literature on perceptions the use o f the term "sex ual offender" occurs again and again. What is not seen is the use of the term "sexual predator" this distinction between the 46

PAGE 47

two groups has thus far been lost in the literature. This is one contribution to the literature that this dissertation aims to fulfill. It is important to examine how the use of these terms effects the community members' perceptions and whether community members fine tune their perceptions depending on whether they are asked about sexual offenders vs. sexual predators. When the word sexual offender is used, do community members only think of sexual offenders or do they think of sexual predators? Or do they equate the two groups together as they think about sex crimes. Predators are legally defined as the mo re severe offender who has a high potential for recidivism ( FDLE 2012 a ) 3 Are all sex offenders, no matter the nature of their crime, lumped together as true folk devils? Community Perceptions of the Sex Offender Registry The Adam Walsh Act ( Public Law 1 0 9 248) has united the sex offender registry throughout the United States and requires that all fifty states comply with the act or forfeit some degree of federal funding. However, although the states are required to comply with the minimum federal standa rds, there is a lot of room for interpretation for state and local governments. This interpretation makes it difficult for community members to get a firm grasp on what a sex offender actually is (Griffin and West, 2006; Logan, 2007). For example in Flor ida, convicted sex offenders are divided into two categories sexual offenders and sexual predators (FDLE, 2012 a ). The state makes a 3 In Florida, sex offenders can be classified as either a sexual predator or a sexual offender. The most basic definition of a sexual predator is someone who has been convict ed of a capital, life or first degree felony sex offense on or after October 1, 1993. In addition, anyone who has any felony violations in addition to the original conviction will be deemed a sexual predator. The court can also deem someone to be a sexua l predator. Also, regardless of meeting these previously mentioned conditions, anyone who has been civilly committed on or after July 1, 2004, must register as a sexual predator (FDLE 2012a ). A sexual offender is defined as someone who has never been des ignated a sexual predator in Florida or in any other state and has committed a sexual offense that is not a capital, life or first degree felony sex offense. Juveniles who have been adjudicated delinquent and who were 14 years of age or older at the time of the crime, can also be designate d as sexual offenders (FDLE 2012a ). 47

PAGE 48

clear distinction between the two groups, but it is unclear if citizens know that difference (Logan, 2007). To make thin gs more complicated, Montana's Sexual or Violent Offender Registration Act (1995) requires that all violent offenders register despite the fact that their crimes are sexual and non sexual in nature. Other states such as New Jersey break up their registr y requirements into tiers, designating sexual offenders as Tier 1 (those "determined to present a low risk of re offense"), Tier 2 (those "found to pose a moderate risk of re offense"), and Tier 3 offenders (those "determined to pose a relatively high risk of re offense") (New Jersey State Police, 2012). Even this short review of three states, demonstrates that there are various styles of registration and community notification. I t is no wonder that community members have a difficult time discerning betwe en offender types but have strong opinions regarding sex offenders in general (Mears et al., 2008; Phillips, 1998; Schiavone and Jeglic, 2009 ). Part of what feeds into a moral panic is the idea that community members have the perception that an event is o ccurring regardless of whether or not it truly is. There is a common misperception that sex crime has been steadily increasing over the last couple of decades. Research has disproved this misconception through analysis of official data from the Uniform Crime Report, which suggests that the amount of forcible rape has steadily decreased since the 1980s (Lonsway and Archambault, 2012; Mancini, 2013). Although there are limitations in using official data (primarily with the dark figure of sex crimes), oth er researchers have examined unofficial data in conjunction with official data sources (Finkelhor and Jones, 2012) and have come to similar conclusions regarding the decrease in sex crimes across all types of abuse. 48

PAGE 49

This misconception about the frequency of sex crime also ties into the misconception regarding the recidivism rates of sex offenders. A national study of community members reported that more than 93 percent of Americans believe this statement about sex offender recidivism to be accurate (Manc ini and Mears, 2010). However, research has set out to debunk that myth. Researchers have suggested that the recidivism level for sex offenders is rather low in comparison to other types of offenders (Sample and Bray, 2003). Over a five year time period it is estimated that only 6.5% of sex offenders will be arrested for a new sex crime (Sample and Bray, 2003). This lead Sample and Bray to conclude, "sex offenders do not appear to be more dangerous than other criminal categories" (2003: 76). Many layp ersons are in favor of the sex offender registry because they perceive it to be an effective tool in monitoring sex offenders living in the community (Pratt, 2000) Even when it has been shown that the registry really has no effect on reducing recidivism rates for sex offenders, there is still a large amount of community support backing the registries (Federoff and Moran, 1997; Pratt, 2000). Overall, most citizens are in favor for more punitive sanctions given to all sex offenders, regardless of the regis try's presence. However, when a child is the victim of a sex offense, there seems to be an even greater desire among the general public to apply a harsh reaction to the child molester compared to other sex offende rs (Mears et al., 2008). Several questio ns still are unanswered. D o the majority of individuals feel this way? Are all citizens as knowledgeable about the registry as they claim to be? Addressing the support issue for the registry, one study found that 4 in 5 residents of Washington State rep orted that 49

PAGE 50

they supported Megan's Law (Phillips, 1998). Despite the large amount of support for the registry notification laws, how accurately do they know the intricacies of the law itself? Prior research conducted on addressing these issues is rather l imited, leaving gaps in the literature. However, in the research that has been conducted, results lean towards high support for the registry but limited or inaccurate knowledge concerning what the registry actually entails (Schiavone and Jeglic, 2009). Other research has been conducted on the opinions of community members regarding the registry and support for its use. Using a telephone survey, Proctor, Badzinski and Johnson examined the role the media plays in influencing knowledge about and support for the sex offender registry and the laws pertaining to registration (2002). Overall, the participants did not demonstrate a high level of knowledge concerning the notifications laws. There was support however for the idea that media coverage does indee d have an influence over the individuals' knowledge levels of the registry and the notification laws. The majority of the respondents supported sex offender registration and believed that it is effective in protecting community me mbers (Proctor et al., 20 02). In one of the earliest studies addressing the issue of support for the registry, Redlich (2001) examined the perceptions of law enforcement officers, law school students and layperson community members in relation to the effectiveness of the registr y. Although there was strong support among the different participant groups, there were some variations in terms of the degree of support. Not surprisingly given their enforcement responsibilities, law enforcement officers supported the registry the most In other words, law enforcement officials are required to serve and protect the 50

PAGE 51

community from offenders, so because of their duties as an officer they would be more likely to want to be punitive towards those individuals who are in opposition to them. L aw school students perhaps reflecting their exposure to adversarial analysis, we re the least supportive and lay person community members fell somewhere in between. Community members would likely include parents, who would be supportive of the registry e fforts as a way to protect their children. Due to their roles within society, participant support may be influenced by their personal roles and biases. In addition to these findings, Redlich also reports that there is an inverse relationship present betw een knowledge about and support for the registry meaning that as knowledge levels increase, support for the registry decreases (2001). In another telephone survey, Lieb and Nunlist surveyed 643 individuals concerning the Washington State registry thi s study addressed opinions about sex offenders and community notification, and the purpose of community notification (2008). Most respondents reported being familiar with the notification laws (82%) and learning about them through some sort of media outle t (63%). As far as safety is concerned, 78% of participants felt safer as a result of sex offender registration. However, even though the registry is useful to communities as a safety tool, it is recognized that sometimes there are unintended consequence s that result from having your personal information made public knowledge (Schiavone and Jeglic, 2009). Reports of vigilantism are not uncommon regarding attacks or harassment of sex offenders, therefore 78% of respondents felt that something should be do ne to make sure such vigilante justice does not occur (Lieb and Nunlist, 2008). 51

PAGE 52

Community Attitudes Towards Sexual Offenders (CATSO) Scale Although the literature surrounding sex offenders and the registry is limited, there have been advancements in how to accurately measure perceptions of these groups. In one of the earliest studies to systematically measure perceptions of sex offenders, Ho gue (1993) used the Attitudes Toward Sexual Offenders Scale (ATS), which was adapted from the Attitudes Toward Prisoners Scale (ATP) (Melvin, Gramling and Gardner, 1985). Hogue's study determined that the police hold the most negative attitudes towards se x offenders but those practitioners who have frequent interactions with sex offenders tend to have more positive attitudes towards sex offenders (1993). In replication research, females tended to view sex offenders less negatively then males. Males also perceived sex offenders with minor victims more negatively then those who committed rape against an adult. Female participants did not make the same distinctions that the male participants did (Ferguson and Ireland, 2006). Additional research has focused on civilian, practitioner and law enforcement perceptions of sex offenders. While the studies have added to the currently growing body of literature, there has been a lack of a systematic sex offender measure. To solve this limitation, Wesley Church and colleagues developed a set of measures that specifically address attitudes towards sex offenders (Church, Wakeman, Miller, Clements and Sun, 2008). Part of the reasoning behind this specific measure comes from the idea that sex offenders are a very stigma tized and target group of offenders, but often citizens are not as knowledgeable about sex offenses as they think they are however, the emotional response is still present (Griffin and West, 2006; Church et al., 2008). 52

PAGE 53

Church and colleagues developed the finalized version of the Community Attitudes Toward Sex Offenders (CATSO) Scale after conducting two pilot studies, which served to concentrate the total number of measures and to verify the reliability of the scale (2008). The initial study began with 3 32 measures, which were cut down to 101 after extensive review by the authors and other sex offender experts. From there an additional four items were eliminated due to clarity issues, leaving 97 items to be tested (Church et al., 2008). The first pilot study used an exploratory factor analysis to test these 97 items on a n undergraduate sample of 347 participants. From this analysis, the researchers decided to remove 20 items due to skewed responses, 47 items due to low factor loadings, leaving 30 items to be tested in the second pilot study (Church et al., 2008). The factor loadings of the remaining 30 items created four groups of measures: 1) Social Isolation (10 items), Capacity to Change (7 items), Severity/Dangerousness (7 items) and Deviance (6 ite ms). The items in these categories will were further cut down (Church et al., 2008). In the second pilot study, Church et al. once again used undergraduate students; this time 316 similar participants were chosen for the study (2008). In examining the 30 measures, the researchers decided that three of the Deviance measures should be removed due to low factor loadings and nine measures were removed because they has "standardized lambdas lower than .40" (Church et al., 2008: 257). This left the researche rs with an 18 item finalized version of their scale. The Church et al. CATSO Scale was developed as a specific systematic way to measure community attitudes regarding sex offenders (2008). The authors advocate its use in many arenas, such as collecting da ta from mental health practitioners, 53

PAGE 54

correctional officials and law enforcement. In addition, the scale will be really useful in gathering information from the general public something that will done in this study. Table 1 1. Finalized CATSO Scale Facto r Loading Measure Social Isolation (Crohnbach's alpha: .80) Sex offenders have difficulty making friends even if they try real hard. Most sex offenders do not have close friends. Most sex offenders keep to themselves. Sex offenders prefer to stay home alone rather than be around lots of people. Most sex offenders are unmarried men. Capacity to Change (Crohnbach's alpha: .80) Convicted sex offenders should never be released from prison. Sex offenders should wear tracking devices so their location can be pinpointed at any time. People who commit sex offenses should lose their civil rights (e.g. voting and privacy). Trying to rehabilitate a sex offender is a waste of time. With support and therapy, someone who committed a sexual offense can learn to change their behavior.* Severity/Dangerousness (Crohnbach' s alpha: .7 0) A sex offense committed against someone the perpetrator knows is less serious then a sex offense committed against a stranger.* Only a few sex offenders are dan gerous.* Someone who uses emotional control when committing a sex offense is not as bad as someone who uses physical control when committing a sex offense.* The prison sentences sex offenders receive are much too long when compared to the sentence leng ths for other crimes.* Male sex offenders should be punished more severely than female sex offenders.* Deviancy (Crohnbach' s alpha: .43 ) Sexual fondling (inappropriate unwarranted touch) is not as bad as rape. Sex offenders have high rates of sexual activity. People who commit sex offenses want to have sex more often then the average person. *These items were reverse coded for analysis purposes. **All items were measured on a six point Likert Scale with response options being: 1) Strongly Disagree, 2) Disagree, 3) Probably Disagree, 4) Probably Agree, 5) Agree, 6) Strongly Agree ***Total CATSO Scale reliability Crohnbach' s alpha: .74 54

PAGE 55

Other researchers have already begun using the CATSO Scale to research attitudes towards sex offenders. In one of the first studies to test the CATSO Scale, Balow and Conley applied the instrument to probation/parole officers and other correctional workers in Montana (2008). A total of 307 individuals participated in the study 4 and were recruited through the Montana Department of Corrections via email. There was some level of "snowballing" that occurred, but the predominant participants were gathered from the recruitment process (Balow and Conley, 2008). The authors only provided frequency distribution bar graphs i n their report, but used the same measures and the same six point Likert Scale to measure the participants' attitudes towards sex offenders (Balow and Conley, 2008; Church et al., 2008). The results of the survey show that practitioners perceive sex offenders as a dangerous population but still believe them to be amenable to treatment (Balow and Conley, 2008). In addition, there was not a lot of support for the idea that sex offenders should never be released from prison or that they should be requir ed to wear tracking devices this is something that the authors attribute to the likelihood that "the real world logistics of this affected opinions" (Balow and Conley, 2008: 6). Although this study was one of the first ones to test the CATSO Scale, ther e are serious limitations in the study. For example, the scale was the only thing that was tested there were no demographic questions asked and no additional sex offender measures asked. Although this is acknowledged by the authors, this lack of additi onal data hurts the integrity of the study and makes further analysis difficult. This report was later published as a peer reviewed 4 Participants included Parole and Probation Officers, Pre Release Center Workers, Administrative Staff, Client Advisors, Case Managers, Licenses Additions Counselors, Intensive Parole and Probation Officers and Registered Nurses (Balow and Conley, 2008). The authors do not describe the occupational duties of all of these Montana Department of Corrections employees, so it is difficult to determine exactly what a client advisor or a lice nses additional counselor is designated to do. 55

PAGE 56

journal article (Conley, Hill, Church, Stoeckel and Allen, 2011) but there were no new elements to the article that were no t reported in the report to the Montana Department of Corrections (Balow and Conley, 2008). The literature surrounding the CATSO scale is very limited, but there is a recent article by Tewksbury, Mustaine and Payne that addresses the viewpoints of criminal justice officials There is a great need to continue to test this scale, to see if it has the potential to become a standard form of perceptions measurement. Up to this point, all of the prior research has focused on some sort of practitioner sample, bu t the current study implements the CATSO scale to see if there is support among a layperson community sample. Fear of Sexual Victimization Given the role of threat in the moral panic framework, it is necessary to complete a short review of fear of sexual victimization. If a person has been a victim of a sex offense, then it stands to reason that this individual might have a higher level of fear regarding further victimization, compared to someone who has never been victimized (Ferraro, 1996 ; Hilinski, 20 09 ). The fear of crime literature also states that in general, women are more fearful of all types of victimization compared to men (Akers et al., 1987; Ferraro, 1996; Liska, Sanchirico and Reed, 1988; Warr, 1984). It is important to note that although w omen are generally more fearful of victimization then men, female rates of victimization are generally much lower with the except of rape and intimate partner violence (Ferraro, 1996 ; Lane, Gover and Dahod, 2009 ). One possible explanation for this inverse relationship is the idea that fear of sexual assault trumps all other fear o f crime for women (Ferraro, 1995; 1996 ). Ferraro hypothesized that "women's fear of rape and sexual assault increases their fear of nonsexual crimes 56

PAGE 57

because any type of victimiza tion could escalate into a rape or sexual assault" (1995; Hilinski, 2009: 85). Ferraro hypothesized that fear of sexual assault overtakes all other types of fear, labeling this the "shadow of sexual assault hypothesis" (Ferraro, 1996, F isher and Sloan, 2 003; Hilinski, 2009; Wilcox et al., 2006). In examining this possibility, Ferraro found support that this might indeed be the case, but only when perceived risk was also taken into account (1996). In addition, there was a positive relationship reported b etween fear of crime and constrained behaviors (meaning that people alter their routine activities, as a protective measure in order to lower their risk of victimization), something that was previously supported in the literature (Liska, Sanchirico and Ree d, 1988; Taylor, Taub and Peterson, 1986). In 2003, Fisher and Sloan applied Ferraro's hypothesis to a sample of college men and women. Their research showed that fear levels for women were higher then those for men, when rape and sexual assault were in cluded in the models. When these two items were not included in the model, the results showed that men and women had similar levels of fear of non sexual crimes (Fisher and Sloan, 2003). In addition, there was some evidence that there was more fear prese nt for crimes committed at night compared to daytime crimes. Fisher and Sloan failed to study the offender victim relationship however, a point that they recognize as one of their biggest study limitations. Other research has tried to expand on Ferraro's (1995; 1996) research as well as the Fisher and Sloan (2003) study. Wilcox, Jordan and Pritchard (2006) included the offender victim relationship as part of their expansion of the Fisher and Sloan study. 57

PAGE 58

Again the shadow of sexual assault hypothesis was supported, even when the offender victim relationship was taken into account (Wilcox et al., 2006). What this means is that women are still highly afraid of sexual assault and crime in general, regardless of who their offender might be. Furthermore, is im portant to provide additional information surrounding fear of crime literature surrounding sexual assault. This is necessary because fear of being sexually victimized might have an influence on how individuals perceive sexual offenders. If an individual is already afraid of being victimized, perhaps that fear is amplified because we are now discussing an already feared group. Victimization should also amplify negative perceptions regarding sexual offenders in applying the fear of crime literature to pe rceptions of sex offenders, research would suggest that if a person is victimized, then he or she would not be as tolerant of sex offenders as a whole (Hilinski, 2009). Given the shadow of sexual assault hypothesis and supportive research, we can expect t hat gender will related to fear of sexual victimization. The research shows that women are more fearful of potential victimization. The shadow of sexual assault hypothesis posits that this will not only extend to sexual victimization, but to other percep tions relevant to sex offenses. It is reasonable to explore how gender relates to fear of personal victimization, fear of victimization of children, and negative attitudes related to sex offenders. Prior Literature on the Stereotypical Sex Offender Additi onally, the idea of the "symbolic assailant" will also be briefly examined (Skolnick, 2007, 2010). First developed by Jerome Skolnick the symbolic assailant is described as almost the stereotypical version of the offender that we have come to associate w ith certain types of crimes. In police work, this comes from the descriptions 58

PAGE 59

given over the call radios (Skolnick, 2010) and responsibility of law enforcement to examine every offender that fits the description given. Applying this idea to sex offenders there could be a case made for the idea that nearly all sex offenders are vi ewed as predators who seek out children. While the idea of the symbolic assailant has not been frequently used in the sex offender literature (Nelson, 1994), it is necessary to find out what image comes to mind when community members think about sex offenders. M edia coverage ha s portrayed sex offenders strictly as predators who are after our children these individuals are just lurking in the shadows waiting to strike. These t hings combine enhances the idea that sex offenders are to be feared and we must take proactive steps to keep our children and ourselves safe. Stereotypes can and have effected a variety of social groups. In the 1950s and 1960s, the media reported on a var iety of misconceptions related to African Americans as they fought for equal rights during the beginnings of the Civil Rights Movement (Allport, 1954/1979). Mass communication researchers have found that the distortion created by the media does its best t o demonize the outsiders who are different from the "favored social elites" (Gorham, 2006: 290; Greenberg, Mastro and Brand, 2002). Although stereotypes have certainly been known portray racial groups a specific way, these stereotypes can be used to creat e a specific caricature based on the specifics related to crime. One finding from stereotyping literature suggests that it is important to learn how audiences process the media information that they are receiving. Even if they are intelligent, logical in dividuals, some ever present media attention can strengthen the stereotype and can increase prejudice or bias in the viewer (Devine, 1989; Monteith, 1993). For example, through the popular culture (TV shows, movies, 59

PAGE 60

and books) there stereotypical sex offe nder has been created. That stereotype does not promote the notion that the most common offender is a younger, white man who knows his victims (see Ackerman et al. 2011). But instead the stereotype is presented as an older, white man who is a stranger to his victims, and will often abduct them to commit his crimes the problem with this stereotype is that many others potentially could avoid detection because they do not fit the mold of what type of sex offender is most threat ening (Sanghara and Wilson, 2006 ). The stereotype about the individual can also extend to the behavior of the group. For example it is a large misconception that sex offenders are chronic recidivists, but research states the opposite that sex offenders have low recidivism levels (Mancini, 2013). It is problematic when the stereotype overrides empirical knowledge, as it then provides fuel to the moral panic fire. The media attention surrounding these issues reinforces the "folk devil" classification that many sex offenders now f ace, while providing support to law enforcement officials charged with supervising and maintaining control over these pariahs. The literature has shown that there is a large amount of support for the sex offender registry throughout communities and negat ive perceptions associated with sex offenders in general Individuals feel that the registry is effective in protecting them from sex offenders, who are perceived as threats to the community as a whole, and to children in particular. Therefore, because o f this large amount of support concerning the registries, there needs to be more research conducted on how well the registries and notification laws are understood. In other words, are people truly knowledgeable about the registry? Or are their answers e motionally driven and given in such a way 60

PAGE 61

because that is what they want to believe? There is not enough literature that presents a comparison between individuals either. When people are concerned about a potential threat, they are more likely to monitor and look out for that threat. Does this have any impact on how the registry is perceived? Do these perceptions differ among people who feel threatened and who do not feel threatened? Since there are so many questions that the literature does not cover in depth, this study proposes a careful examination of some of these issues specifically acknowledging the link between support for the registry and knowledge levels of the individuals. Two groups of community members one comprised purely of parents a nd the other a mixed sample of parents, people without children and grandparents will serve as a direct comparison to examine this previously discussed relationship. The following sections of the paper will propose the intended research study and how th e data will be collected. 61

PAGE 62

CHAPTER 3 RESEARCH METHODOLOGY This dissertation is topic focused rather than being derived from theory or arising from tensions in the research literature. Its topic encompasses a constellation of features surrounding sex offenses and the registry laws that are in place in the United States. Its primary purpose is to explore those features using survey data from a national sample to move toward a more complete and cohesive understanding of our reactions to sex offenses an d offenders The dissertation uses the literature on "moral panics" as a sensitizing theoretical framework to help organize the research. The features of a moral panic include: deep concern over the problem, hostility or anger toward the perpetrators, broa d consensus about the issue, a call for severe reactions that are disproportionate to the empirical evidence on the problem, and volatility in our reactions. The dissertation will explore at the general level the consensus themes suggested by "moral panic s." General hypotheses are derived in accordance with those expectations. One of those expectations is that we should not see important differences in citizens' responses by most social categories or groupings like age or race or education or marital sta tus or family income. Despite its consensus underpinnings, the moral panic approach is not as singular as it might seem at first glance. In fact, there are some features of applying the moral panic lens to this topic that have implications for hypothese s and research methods. As with any social problem that leads to calls for action, some groups will be more involved and more activated by the perceived problem. We expect that apply to the problem of sex offenses and offenders. As developed below, the dissertation breaks out a few 62

PAGE 63

exceptions to the general consensus expectations and advances specific exploratory hypotheses as a way of trying to get at the subtleties of this topic. It also incorporates an experimental manipulation to delve deeper into t he phenomenon. A moral panic involves hostility and anger toward the offense and offender. The presence of a "folk devil" points to the need to consider the "demonizing" features of sex offender stereotypes. Two often inter related features are sugges ted immediately: child victims and sex predators. This prospect is incorporated into the research in several ways. One implication is raised by the potential concern about child victims. It puts the spot light on parents who may be more sensitized and attendant to sex offenses and offenders than are other citizens. We may expect parents to have a different level of perception and involvement with the perceived threat of sex offenses and offenders. Accordingly, specific hypotheses will be derived to compare responses from parents with those from nonparents. Another research implication of the demonization deals with the distinction, made in law, between "sex offenders" and "sex predators." The research utilizes two different survey forms, one of w hich refers to "sex offenders" and the other of which refers to "sex predators" for many of the questions. Those forms are randomly assigned to respondents. This allows the research to explore whether "folk" or "popular" knowledge stereotypes make a dist inction and whether the manipulation in terminologies systematically affects responses on questions. The sex offender registry laws themselves (which are at least severe reactions if not disproportionate ones) are a springboard into the investigation. One focus of the 63

PAGE 64

research problem concerns how often citizens access the registries and how much "legal" knowledge they have about the registries. That knowledge and access may very well relate to perceptions that citizens have about sex offenses and offende rs the kind of "folk" or "popular" knowledge individuals rely on when contemplating sex crimes and criminals. Do citizens view a kind of "folk demon" that becomes the basis for social control efforts? For this reason, measures of general stereotypes of o ffenders/offenses are also included with measures of registry knowledge and accessing behavior in a research area focusing on citizens' "personal orientations toward the control of sex offenders." The misleading nature of stereotypes is one way to conside r the disproportionality of severe legal reactions like the registry. The source of the hostility toward the "folk devil" in moral panics (and a unifying factor in criminalization efforts) may well be threat or fear of victimization. Accordingly, the second research area examines "fear of sexual victimization." Because the child victim is the more "demonizing" feature, the research examines citizens' fears of personal victimization for themselves and their fears about the victimization of children. Con sistent with the "shallow of sexual assault" hypothesis, the research expects gender differences in fear of sexual victimization and explores for that prospect for other measures regarding citizen orientations to sex crime phenomena. A moral panic must go beyond individualistic concerns like fear of victimization or levels of knowledge or behaviors like accessing the registry. The panic implicates the community. The third area of focus in this dissertation research is on the "perceptions of community rela ted attitudes," the most obvious of which would reflect the dimensions of moral panic: deep concern over the problem (itself a potential dimension of fear of 64

PAGE 65

crime, not always highly related to victimization of self or others), hostility or anger toward th e perpetrators, broad consensus about the issue, a call for severe reactions that are disproportionate to the empirical knowledge of the problem, and volatility in our reactions. This research will explore the extent of concern, the hostility shown, the d egree of consensus, and the disproportionality of reaction. It will attempt to tap into volatility as well. However, one of the central players in the moral panic literature is more cautious about the volatility feature. Cohen argues, "sometimes the pan ic passes over and is forgotten, except in folklore and collective memory; at other times it has more serious and long lasting repercussions and might produce changes" (Cohen 2004:1). The dissertation assumes that our reactions to sex offenses and offende rs are being sustained, have produced changes, and await the next tragedy to energize the call for even more changes. Some of those changes can be seen in the series of "memorial criminal justice policies" (see Surette, 2011:2) that capture our "panic" wh en yet another tragedy occurs (e.g., Megan's Law, Amber Alerts, Adam Walsh Child Protection and Safety Act, etc.). Indeed, a recent Florida kidnapping, molestation and murder of a young child has prompted a new round of legislative proposals to increase s everity of sanctions in that state (Pan z at i & Treen, 2013). Another way to consider community related attitudes exists in the sex offense literature. Accordingly the dissertation research employs measures on "community attitudes towards sex offenders" ( or the CATSO scale and its sub dimensions of isolation, capacity to change, severity/dangerousness and deviancy). One last community related theme ties back to the original research impetus stemming from 65

PAGE 66

registries. The dissertation examines the level of support that citizens show for the registry itself. Exploratory Hypothes es Specifically, this research project is seeking to determine if there is a moral panic occurring in regards to sex offenders, how fearful community members are of sex offenders l iving among them and the levels of knowledge they equate with the sex offender registry. Therefore, it would be expected that some community members are going to have higher levels of fear and panic associated with sex offenders rather than others. Thes e hypotheses also address some of the post panic issues that have been previously discussed specifically accessing the registry and the levels of fear could be attributed to this post panic legal atmosphere that we find ourselves in. The Community Attit udes Toward Sex Offenders (CATSO) Scale also is highlighted in this post panic point of view. Furthermore, the randomization between the terms "sex offenders" and "sexual predators" strengthens the previous literature surrounding perceptions of sex offend ers, specifically in regards to researchers not using the term "sexual predator" in previous studies. Based on the previous literature and the project goals, several hypotheses have been developed to test these propositions. These hypotheses start from a general perspective regarding attitudes towards sex crimes, offenders and the registry, and then narrow down to predict differences between parent and non parent participants. A series of hypotheses are advanced to explore the constellation of relationshi ps surrounding citizens' orientations regarding sex crimes/criminals and the law. The hypotheses are grouped into the three areas: personal orientations towards the control of sex offenders, fear of sexual victimization, and community related attitudes. More 66

PAGE 67

specific measurements subsumed under each area are then linked to hypotheses that move from general descriptive expectations to highlighting expectations for some specific relationships between and among variables. Exploratory Hypotheses Concerning the "Personal Orientations Toward the Control of Sexual Offenders." 1. Accessing the Registries (Randomized Offender/Predator Manipulation) a) Generally, relatively few people will access the registries. b) More specifically, the extent to which people access the registries will be similar across social categories and groups c) The exceptions to b are: i) is that parents of children will access the registries more often than will non parents ii) that women will access the registries more often than will men. 2. Knowledge abou t the Registries. (No Randomized Offender/Predator Manipulation) a) Generally, the level of knowledge about registries will be low. b) More specifically, the level of knowledge will be similar across social categories and groups c) The exception to b are: i) That pa rents of children will have higher knowledge than non parents. ii) That women will have higher knowledge than men. 3. "Popular" knowledge/stereotype about sex offenders and sex predators ( No Randomized Offender/Predator Manipulation) a) Generally, the level of ina ccura te stereotyping will be high. b) More specifically, the level of inaccurate stereotyping will be similar across social categories and groups. c) The exceptions to b are: i) Those with more knowledge about the registries will have lower levels of inaccurate ste reotyping ii) P arents of children will have higher levels of inaccurate ster eotyping than will non parents. iii) Women will have higher levels of inaccurate stereotyping than will men. Exploratory Hypothesis About Fear of Crime ( No Randomized Offender/Predator Manipulation) 4. Fear of Personal Victimization a) Generally, the level of fear of personal victimization will be moderate. b) More specifically, the level of fear of personal victimization will be similar across social categories and groups. c) The exception to b are : 67

PAGE 68

i) That non parents will report more fear of personal victimization than will parents. ii) That women will report more fear of personal victimization than will men. 5. Fear of Victimization of Children a) Generally, the level of fear of victimization of children will be high. b) More specifically, the level of fear of personal victimization will be similar across social categories and groups. c) The exception to b are: i) That parents will report more fear of victimization of children than will parents. ii) That women will report more fear of victimization of children than will men. Exploratory Hypotheses about "Perceptions of Community Related Attitudes" Citizen endorsement of Moral Panic Features (Randomized Offender/Predator Manipulation) 6. The Concern Dimension a) Generally, the con cern about sex offenses and offending will be high among the citizens. b) More specifically, that concern will be similar across social categories and groups and it will be similar for the group responding to "sex offender" and that responding to "sex predat or" survey items c) The exceptions to b are: i) T hat parents will be more concerned than nonparents ii) T hat women will be more concerned than men. 7. The Hostility/Anger Dimension a) Generally, the hostility/anger about sex offenses and offending will be high among the citizens. b) More specifically, that hostility/anger will be similar across social categories and groups and it will be similar for the group responding to "sex offender" and that responding to "sex predator" su rvey items c) The exceptions to b are: i) T hat parents will be more hostile/angry than nonparents. ii) That women will be more hostile/angry than men. 8. Generally, the consensus about sex offenses and offending will be high among the citizens. a) More specifically, that consensus will extend across social categories and groups and it will be similar for the group responding to "sex offender" and that responding to "sex predator" survey items b) The exceptions to be are: i) That parents will be more unified than nonparents. ii) T hat women will be more unified than men. 9. The Disproportionate Reaction Dimension a) Generally, the reaction about sex offenses and offending will be more disproportionate among the citizens. 68

PAGE 69

b) More specifically, that disproportionate reactions will extend acros s social categories and groups and it will be similar for the group responding to "sex offender" and that responding to "sex predator" survey items c) The exception to b are: i) T hat parents will react more disproportionately than will nonparents. ii) That women w ill react more disproportionately than will men. 10. The Volatility Dimension to be explored, no a priori hypotheses advanced 11. Community Attitudes Towards Sex Offenders using the CATSO Scale (No Randomized Offender/Predator Manipulation) a) Generally, citizens w ill have highly negative attitudes for the overall scale and on each of the four sub dimensions (social isolation, capacity for change, severity/dangerousness, and deviancy). b) More specifically, these attitudes will be similar across social categories and g roups. c) The exceptions to b are: i) Those with registry knowledge will hold more negative attitudes on the CATSO scale and sub dimensions than will those with less registry knowledge. ii) Parents will hold more negative attitudes on the CATSO scale and sub dimensions than will nonparents. iii) Women will hold more negative attitudes on the CATSO scale and sub dimensions than will men. 12. Community Member Support for the Sex Offender Registry (Randomized Offender/Predator Manipulation) a) Generally, citizens will s how high support for the registry, and that support will be similar for the group responding to "sex offenders" and for the group responding to "sex predators." b) More specifically, the support will be similar across social categories and groups for both the group responding to "sex offenders" and that responding to "sex predators." c) The exception to b are: i) T hat parents will be more supportive of the registry than will nonparents. ii) That women will be more supportive of the registry than will men. Participants This dissertation results from pilot research and a full study, both of which were approved by the university's Institutional Review Board. The pilot was conducted with college students from a participant pool at a large southeastern university. The full study recruited adult community members from a national website. The pilot 's 69

PAGE 70

participants access ed a survey on a participant pool website which is run through the criminology department at that university. The pilot study was conducted before the communit y members were recruited to try out the survey instrument, to check out the operationalizations, and to work on the scaling for multiple item measures of important constructs. This allowed for adjustments before posting the survey for the adult community members in the full study. Recruitment of Participants Pilot Participants The participants for the pilot were undergraduate students enrolled in criminology classes at a large southeastern university which were p articipating in a student participant poo l at the university Only students over 18 years of age without children were asked to participate. Students received one unit of research credit for their participation. A table in Appendix B ( Table B 1 ) shows the participant demographics for the pilot study. Full Study Participants After the pilot study was completed, recruitment of the community member sample began. Participants were recruited nationwide from an online participant pool conducted via Mechanical Turk (MTurk) on the Amazon.com website. MTurk is described as an online marketplace where employers (researchers) can hire employees (participants) to take part in a variety of work. Participants are compensated for their time based on whatever amount the researcher is willing to pay. For thi s project, participants were offered a $1 incentive for taking the survey, which was rather high in comparison to the other projects being offered the day of collection. MTurk required that the researcher prepay into an account before recruitment can begi n. 70

PAGE 71

A n advertisement was posted on the MTurk website inviting participants to take part in a criminological survey pertaining to perceptions of sexual offenders. Participants were told about the $1 incentive and that the survey would take them between 30 m inutes to an hour. A link to the survey was listed in the invitation directing the participants right to the survey website. Because the researcher was required to prepay the MTurk account, there was no need to collect any personal information for paymen t purposes. Amazon.com served as third party and paid the participants from the prepaid account set up by the researcher. Therefore, the answers remained anonymous as the participants' contact information was not collected. A total of 877 surveys were s ubmitted. A breakdown of the full study's sample is provided in the Results chapter (see Table 4 1 below). Instrumentation Pilot Instrument The survey instrument for the pilot study focused on registry knowledge (used as the independent variable in the full study) and fear levels of respondents. The pilot study also operationalized the five elements of a moral panic concern, hostility/anger, unification/consensus, disproportionality, and volatility. There were six questions used as control variables, asking participants their gender, age, race, ethnicity, academic class standing and whether or not they were victimized of a crime. The main components of the pilot the registry knowledge variable and the five elements of the moral panic were analyzed to see if the measures were accurately measuring the constructs of the study. The pilot did not manipulate "sex offender" or "sex predator" terms on different forms of the survey instrument. The pilot study instrument, frequency tables and preliminary data analysis are all located in Appendix A of the paper 71

PAGE 72

Ten items addressed registry knowledge. Five questions were included to t ap each of the five moral panic elements, for a total of twenty five moral panic items. Finally, the Community Attitudes Towards Sex Offenders (CATSO) Scale (see Church et a. 2008) was added as a way to gauge attitudes held towards sex offenders. The pil ot study measures were based on Florida laws. The following tables show the factor analyses conducted. Table 3 1. Factor Analysis for Registry Knowledge Component 1 2 3 4 Q15 Everyone who has ever been convicted of a sexual offense is required to register on the Florida Sex Offender Registry. .499 Q16 The definition of a sexual predator is: Repeat sexual offenders, sexual offenders who use physical violence, and sexual offenders who prey on children are sexual predators who present an extreme threat to the public safety. Sexual offenders are extremely likely to use physical violence and to repeat their offenses, and most sexual offenders commit many offenses, have many more victims than are ever reported, and are prosecuted for only a fraction of their crimes. .769 Q17 Individuals convicted of their very first sexual crime can be classified as a sexual predator on the Florida Sex Offender Registry. .668 Q18 Juvenile offenders can be required to register on the Florida Sex Offender Registry if convicted of a sexual offense. .747 Q19 After serving their prison sentences, sex offenders can be incarcerated indefinitely through the process called "Civil Commitment." .655 Q20 In Florida, there are more male sex offenders registered than female sex offenders. .759 Q21 Female sex offenders make up roughly one fourth of the total amount of registered sex offenders. .766 Q22 The Amber Alert System is named after the color amber, at no time was the system designed in memory of a child. .330 Q23 Sex offenders have one of the highest recidivism or re offending rates of any offenders. .448 Crohnbach's alpha = .373 Table 3 1 presents the factor analysis for the nine measures of registry knowledge. They did not coalesce well in that they loaded across four different components and combining them into a single measure yielded a low reliability estimate (Cronbach's alp ha=.373). The results prompted some adjustments for the instrument 72

PAGE 73

that would be used for the full study: pilot survey questions 15, 16, and 21 would be removed; questions 17 and 23 would be reworded; and several new items would be added. Table 3 2 prese nts the results for the Moral Panic items. The factor analysis shows that the 23 survey items load on five elements. This is the number of components expected for the five elements of a moral panic, but there were a few issues with several of the items th at prompted adjustments for the full study instrumentation. The expected items loaded well for the Concern, Hostility, and Consensus elements, but the factor analysis results show that Disproportionality and Volatility had the least consistent loadings of the five moral panic elements. These may be the most difficult to measure in a cross sectional study. For the full study, one item would be added to both the concern and hostility scales. Additionally, several of the items would be reworded to adapt th e instrument for a community wide sample instead of a college student sample. These changes were effective in overcoming the problems with the initial factor loadings (see Table 4 21 for the full study factor analysis). Once the surveys were completed, an invitation was issued for 20 percent of the sample to be interviewed further on the relevant information concerning the sex offender registry and the moral panics issues 5 In order to be eligible for the interview, the participant had to have taken the q uantitative survey before they could participate in the qualitative interview. The interview asked the participants to elaborate on their answers from the survey specifically in the areas of registry knowledge, the level panic regarding sex offenders li ving the community and personal fear being a victim of a sex 5 The goal of the pilot study is to recruit roughly 400 participants. Therefore, to meet the 20 percent interview goal, 80 interviews will need to be conducted. 73

PAGE 74

Table 3 2. Factor Analysis for Moral Panic Measures Component 1 2 3 4 5 Q24 Are you worried about sex offenders living nearby your home? (Concern) .717 Q25 Are you worried that you personally may be a victim of a sexual offense? (Concern) .785 Q26 Are you concerned about sex offenders being on the college campus? (Concern) .765 Q27 Are you worried that if sex offenders are living in the community, then more sexual offenses will occur? (Concern) .588 Q28 Are you angry that sex offenders are allowed to live in the community? (Hos t ility) .790 Q29 Do you feel any resentment over the fact that some of your neighbors may be sex offenders? (Hos t ility) .682 Q30 Do you feel any anger towards the criminal justice system for releasing sex offenders from jails and prisons? (Hos t ility) .776 Q31 Are you angry that sex offenders may be working at businesses where you may frequently shop or visit? (Hos t ility) .795 Q32 Do you feel that a majority of community members are in agreement about the risk that sex offenders pose? (Consensus) .455 Q33 Do you feel that many community members feel that changes must be made in the supervision of sex offenders? (Consensus) .609 Q34 Do you feel that community members in general feel threatened by sex offenders as a group? (Consensus) .761 Q35 Do you feel that a majority of community members are in agreement that children are at ris k of being sexually victimized? (Consensus) .788 Q36 Do you feel that many parents feel that sex offenders are too dangerous to be living in the community? (Consensus) .758 Q37 Do you feel that the current state of the sex offender registry is too harsh? (Disproportionality) .668 Q38 Do you feel that the sex offender registry laws should be stricter? (Disproportionality) .660 Q39 Do you feel that keeping sex offenders on electronic monitoring/GPS tracking for more than 5 years without a break is too severe a punishment? (Disproportionality) .707 Q40 Do you feel that sex offenders should report to law enforcement more than the required two times per year? (Disproportionality) .447 Q41 Do you feel that the media overreacts in their reporting of sex offenses when they occur in a community? (Disproportionality) .729 Q42 Do you feel that law enforcement reacts quickly enough when a sexual offense takes place? (Volatility) .792 74

PAGE 75

Table 3 2. Continued Component Q43 Do you feel that legislators work fast enough to get necessary registry laws passed to further keep track of sex offenders? (Volatility) .763 Q44 Do you feel that the media reports sex offenses cases too quickly before all of the facts are gathered? (Volatility) .668 Q45 Do you feel that the quick response of the media makes communities safer because people are made aware of the sex offense? (Volatility) .280 Q46 Do you feel that police are too slow to catch sex offenders when sexual offenses take place? (Volatility) .367 offense. The interviews were not very long in duration but rather they covered the more important aspects of the quantitative survey. Both the quantitative and qualitative instruments, as well as preliminary data findings are listed in the appendix of this paper. Full Study Operationalization Full Study Instrument Only a few changes have been made in the instrument of the full study. The registry knowledge portion of the paper was finalized and a few new measures were added. Additional control variables were added such as number of children, a revised race measu re, information about whether the participant is a grandparent and about the children/grandchildren of the participants. The alternative forms for the full study also converted language about the registries to reflect the relevant federal laws rather than Florida's law. The final survey items for the full study was very similar to the one used in the pilot study The most important instrumentation change for the full study is the experimental manipulation of the terms "sexual offender" and "sexual pred ator" via the random assignment of two alternative forms of the instrument to respondents. Respondents either received a form in which the questions consistently referred to "sexual offender" 75

PAGE 76

or the alternate form where those same questions referred to "s exual predator." This manipulation allows the comparison of responses to see how much community members distinguish between types of offenders. Although there was random assignment, there were more usable surveys from the "sexual offender" assignment com pared to the "sexual predator" assignment. This will be discussed in more detail later. The purpose of the research is to examine a constellation of variables relevant to sex crimes, sexual offenders, and the offender registries to reach a better understa nding of community members' reactions. The variables of interest are grouped into three broad categories: personal orientations toward the control of sex offenders, fear of sexual victimization, and perceptions of community related attitudes. As the hypot heses indicate, for some purposes the same variable may serve as an independent variable but for other analyses it may become the dependent variable. The respective operationalizations are discussed in this section. Personal Orientations toward the Contr ol of Sex Offenders Accessing the Registry Participants were asked about their experience in accessing the sex offender registry in their states. This was done by asking the following questions: 1) "Have you ever looked at your state's sex offender regis try website?" (response options: No, Yes) 2) "Have you ever searched your state's registry website for sexual offenders (predators) living in the areas surrounding your home? (response options: No, Yes) and 3) "How many sex offenders (predators) would you estimate are living nearby your home?" (Only shown if participants answered "Yes" to the second measure response options: Drop down tab of 0 100 sex offenders/predators). 76

PAGE 77

Only the second and third measures were randomized to ask about sexual offenders and predators. These measures will not be included in any regression models, but they will be tested to see if there are any significant differences between participants across social groups. Registry Knowledge The second variable in this category combined te n items into a scale measuring Registry Knowledge. These items were presented without distinguishing between "sexual offenders" and "sexual predators" since that is a legal distinction that is part of the ba ttery of knowledge questions. The Registry Knowledge variable was operationalized through the same ten questions for both the parent and non parent participant group s: 1) In my state, all sex offenders are classified the same no matter their crime 2) Registered sex offenders are required to live at least 1,000 feet from a school zone, park or bus stop" 3) "Some sex offenders ar e required to register for life 4) "Juvenile offenders, who are 14 years old at the time of the offense, can be placed on the registry if convicted." 5) "All sex offenders are required to be on some sort of electronic monitoring/GPS tracking device at all times." 6) "Sex offenders have very high rates of reoffending." 7) "The Amber Alert System is named after a child named Amber, it has nothing to do with the color amber." 8) "T here are more male sex offenders registered than female sex offenders." 9) "Individuals convicted of their very first sex crime can be classifi ed as sexual predator s or can be placed in a Tier III classification 10) "After serving their prison sentences, sex offenders can be incarcerated indefinit ely through the process called Civil Commitment." Response options for all ten questions were "Very True," "Somewhat True," "Unsure," "Somewhat False," and "Very False." Three measures were reverse coded, so these response 77

PAGE 78

options were flipped. These questions were developed primarily for this study and were not reproduced from any previous research. There is no randomization between the terms "sex offender" and "sexual predator" here because the questions are testing the participants' accuracy of knowledge only. These measures are not intended to test the distortion element in perceptions of sex off enders. These ten measures were ultimately transformed into a count variable. Once the data was collected, the five point response options were dichotomized into "incorrect" and "correct" respons e options. This means that answer choices "Very True" and Somewhat True" were collapsed into a correct answer category and the remaining three options "Unsure," "Somewhat False" and "Very False" collapsed into an incorrect answer category. The same procedure occurred for the three reverse coded measures. On ce the responses became dichotomized, the ten measures were transformed into a count variable, producing a range of 0 10. Zero indicates that the participant did not produce any correct responses and 10 means that the participant produced 10 correct respo nses This independent variable will be used in an OLS Regression, tested against the six dependent variables of Fear, Concern, Hostility/Anger, Consensus, Disproportionality and Volatility. The analytic plan to utilize this independent variable is explain ed in Tables 3 4 3 12 or Models 3 10 as seen at the end of the operationalization portion of this paper. Stereotypical Sex Offender The third personal orientation in this category was dealt with "folk" or "popular" knowledge of sex criminals; it looked at inaccurate stereotypical perceptions that respondents attribute to sex criminals. To test participants' knowledge regarding the 78

PAGE 79

most frequent type of registered sex offender, participants were asked to answer seven measures regarding demographic charac teristics of these registrants. Participants were asked to supply responses to the following fill in the blank measures, 1) "Most sex offenders are of the _____ gender (response options: male or female), 2) and are in the _____ age group" (response op tions: 14 25 years old, 26 35 years old, 36 45 years old, 46 55 years old, 56 65 years old, 66 75 years old), 3) "Most sex offenders of the _____ race (response options: Native American/Alaskan, Asian, Native Hawaiian or Other Pacific Islander, Black/Afr ican American, White, Other), 4) and are of the _____ ethnicity" (response options: Hispanic, Not Hispanic), 5) "Most sex offenders are _____ to their victims (response options: Offender is a stranger, Offender is a close friend, Offender is a distant relative, Offender is an immediate relative), 6) and frequently have ______ as their victims" (response options: Pre pubescent female minors, Pre pubescent male minors, Post pubescent female minors, Post pubescent male minors, Adult females, Adult male s), and 7) "_____ is the most frequent form of victimization that sex offenders commit" (response options: Physical non consensual sex act, Physical consensual sex act with a minor, Non physical sex act). Prior literature has suggested that the most freque nt registered sex offender is male (97.7% of the total registered sex offenders nationwide), white (67.0%), has a mean age of 44.3 years of age, had a female vict im (87.0%) between the ages of 11 14 ( 37 .0%), and committed an offense against a child (55.0%) (Ackerman, Harris, Levenson & Zgoba, 2011) The offender/predator randomization was not implemented into these measures. The measures will then be dichotomized into right and wrong responses, based on the 79

PAGE 80

profile developed by Ackerman et al. (2011). The newly dichotomized responses will then be transformed into a count variable, which adds together correct responses. The transformed variable will have a range of 0 7, which indicates the accuracy of knowledge regarding the stereotypical measures; 0 indica tes that the participants did not identify any correct responses and 7 indicates that the participant identified all 7 correct responses. Like the Registry Knowledge variable, the Stereotypical Sex Offender variable will be included in all of the OLS Regr ession models as an independent variable. Fear of Sexual Victimization The first dependent variable developed to test out registry knowledge is the level of fear present in community members. Fear as a dependent variable, is important to test, because man y community members perceive sex offenders as a risk to their personal safety and to the safety of their children. The fear variable was operationalized through t wo questions for the parent and non parent participant groups : 1) "Are you worried that you personally may be the victim of a sexual offense?" and 2 ) "Are you concerned that children in your community may be at risk of becoming victims of a sexual offense?" Response options for both questions were Definitely Not," "Probably Not," "Probably Yes," and "Definitely Yes ." These questions were developed primarily for this study and were not reproduced from any previous research. This dependent variable will be used in an OLS Regression, tested against the independent variable of Registry Knowledg e. The analytic plan to utilize this dependent variable is explained in Table 3 5 Model 3 as seen at the end of the operationalization portion of this paper. 80

PAGE 81

Perceptions of Community Related Attitudes Concern Based on the moral panics literature, five c ategories of dependent variables have been created to see if a moral panic exists in regard to the sex offender registry. The first variable category created from the moral panics literature is concern over sex offenders living in the community. These me asures use randomization of the terms "sex ual offender" and "sexual predator." The concern variable was operationalized through five questions for both the parent and non parent participant group s : 1) "Are you worried about sex offenders (predators) living nearby your home?" 2) "Are you worried that you personally may become a victim of a sexual offense?" 3) "Are you concerned that children in your community may be at risk of become victims of a sexual offense?" 4) "Are you worried about children in your co mmunity being at risk of being approached by a sexual offender (predator) ?" and 5) "Are you worried that a s sex offenders (predators) continue to live in the community, then more sex offenses will occur?" Response options for all five questions were Defin itely Not," "Probably Not," "Probably Yes," and "Definitely Yes ." These questions were developed primarily for this study and were not reproduced from any previous research. This dependent variable will be used in an OLS Regression, tested against the in dependent variable of Registry Knowledge. The analytic plan to utilize this dependent variable is explained in Table 3 6 Model 4 as seen at the end of the operationalization portion of this paper. Hostility The second variable category created from the moral panics literature is hostility and anger over sex offenders living in the community. These measures use 81

PAGE 82

randomization of the terms "sex offender" and "sexual predator." The hostility/anger variable was operationalized for both the parent and non pa rent groups, through five separate questions: 1) "Are you angry that sex offenders (predators) are allowed to live in the community?", 2) "Do you feel any resentment over the fact that some of your neighbors may be sex offenders (predators) ?", 3) "Do you fe el any anger towards the criminal justice system for releasing sex offenders (predators) from jails and prisons?", 4) "Are you angry that sex offenders (predators) may be working at businesses where you may frequently shop or visit?" and 5) "Are you angry that your children might come into contact with a sex offender (predator) ?" Response options for all five questions were Definitely Not," "Probably Not," "Probably Yes," and "Definitely Yes ." These questions were developed primarily for this study and we re not reproduced from any previous research. This dependent variable will be used in an OLS Regression, tested against the independent variable of Registry Knowledge. The analytic plan to utilize this dependent variable is explained in Table 3 7 Model 5 as seen at the end of the operationalization portion of this paper. Consensus The third variable category created from the moral panics literature is unification and consensus over sex offenders living in the community. These measures use randomization of the terms "sex ual offender" and "sexual predator." The unification/consensus variable was operationalized through five separate questions: 1) "Do you feel that a majority of community members are in agreement about the risk that sex offenders (predato rs) pose?", 2) "Do you feel that many community members feel that changes must be made in the supervision of sex offenders (predators) ?", 3) "Do you feel that community members in general feel threatened by sex offenders 82

PAGE 83

(predators) as a group?", 4) "Do yo u feel that a majority of community members are in agreement that children are at risk of being sexually victimized?", 5) Do you feel that many community members feel that sex offenders (predators) are too dangerous to be living in the community?" Respons e options for all five questions were Definitely Not," "Probably Not," "Probably Yes," and "Definitely Yes ." These questions were developed primarily for this study and were not reproduced from any previous research. Response options for all five questio ns were Definitely Not," "Probably Not," "Probably Yes," and "Definitely Yes ." These questions were developed primarily for this study and were not reproduced from any previous research. This dependent variable will be used in an OLS Regression, tested against the independent variable of Registry Knowledge. The analytic plan to utilize this dependent variable is explained in Table 3 8 Model 6 as seen at the end of the operationalization portion of this paper. Disproportionality The fourth variable cat egory created from the moral panics literature is a feeling of disproportionately harsh punishment in regards to the sex offender registry. These measures use randomization of the terms "sex ual offender" and "sexual predator." In order to measure disprop ortionality, it was necessary to first ask two anchor questions to measure the attitudes of participants regarding the registry and the current sex offender laws. The first question asks the participant, "Do you support the use of the publicly available s ex offender r egistry in your state ?" The response options for this question included, "Definitely Not," "Probably Not," "Probably Yes," and "Definitely Yes." The second question asked participants, "How strict do you think the current laws regarding the s ex offender registry ?" The response options for this question were 83

PAGE 84

measured on a five point Likert scale including, "Way Too Lax," "Lax," "Just Right," "Strict," and "Way Too Strict." These two anchor questions were necessary to include because their feelings regarding the registry might have an effect on how they answer the questions regarding disproportionality. If the participants do not support the registry and feel that the curren t laws are too strict, then their responses might indicate that a disproportionate response to sex offenses is occurring that communities as a whole are overreacting to sex offenses and as a result a harsh and strict legal culture has been implemented in reaction. If participants do support the registry and believe that laws are too lax, then the reverse relationship might be true that there is not a disproportionate response taking place in regards to sex offenses and that more could be done to superv ise these individuals, by making the laws stricter and increasing punishments. The disproportionality variable was operationalized for the parent and non parent participant group s through five separate questions: 1) "Do you feel that the current state of t he sex offender registry is too harsh?", 2) "Do you feel that the sex offender registry laws should be stricter?", 3) "Do you feel that keeping sex offenders (predators) on electronic monitoring/GPS tracking for more than 5 years without a break is too sev ere a punishment?", 4) "Do you feel that sex offenders (predators) should report to law enforcement more than the required two times per year?", and 5) Do you feel that the media overreacts in their reporting of sex offenses when they occur in a community ?" Response options for all five questions were Definitely Not," "Probably Not," "Probably Yes," and "Definitely Yes ." These questions were developed primarily for this study and were not reproduced from any previous research. This dependent variable 84

PAGE 85

will be used in an OLS Regression, tested against the independent variable of Registry Knowledge. The analytic plan to utilize this dependent variable is explained in Table 3 9 Model 7 as seen at the end of the operationalization portion of this paper. Volatility The fifth and final variable category created from the moral panics literature is a volatile, quick response in regards to the sex crimes that occur. These measures use randomization of the terms "sex offender" and "sexual predator." The volat ility variable was operationalized for the parent and non parent participant group s through five separate questions: 1) "Do you feel that law enforcement reacts quickly enough when a sexual offense takes place?", 2) "Do you feel that legislators work fast enough to get necessary registry laws passed to further keep track of sex offenders (predators) ?", 3) "Do you feel that the media reports sex offenses cases too quickly before all of the facts are gathered?", 4) "Do you feel that the quick response of the media makes communities safer because people are made aware of the sex offense?", and 5) "Do you feel that police are too slow to catch sex offenders (predators) when sexual offenses take place?" Response options for all five questions were Definitely N ot," "Probably Not," "Probably Yes," and "Definitely Yes ." These questions were developed primarily for this study and were not reproduced from any previous research This dependent variable will be used in an OLS Regression, tested against the independe nt variable of Registry Knowledge. The analytic plan to utilize this dependent variable is explained in Table 3 10 Model 8, as seen at the end of the operationalization portion of this paper. 85

PAGE 86

C ommunity A ttitudes T oward S ex O ffenders (CATSO) Scale In add ition to the moral panic variables, the Community Attitudes Toward Sex Offenders (CATSO) Scale was included as a way to examine individual attitudes towards sex offenders. Wesley Church and colleagues developed the CATSO scale in 2008 after pilot study re search and exploratory factor analysis. The final scale is comprised of 18 items and has been used in several studies (see the review of the CATSO scale in the literature review). For this scale, there is no randomization between "sex offenders" and "sex ual predators." This keeps the scale true to form and does not deviate from the original measures used by Church and colleagues (2008). The scale is operationalized through the following statements: 1) With support and therapy, someone who committed a se xual offense can learn to change their behavior. 2) People who commit sex offenses should lose their civil rights (e.g. voting and privacy). 3) People who commit sex offenses want to have sex more often than the average person. 4) Male sex offenders should be punished more severely than female sex offenders. 5) A lot of sex offenders use their victims to create pornography. 6) Sex offenders prefer to stay home alone rather than be around lots of people. 7) Most sex offenders do not have close fr iends. 8) Sex offenders have difficulty making friends even if they try real hard. 9) The prison sentences sex offenders receive are much too long when compared to the sentence lengths for other crimes. 10) Sex offenders have high rates of sexual activity. 11) Trying to rehabilitate a sex offender is a waste of time. 12) Sex offenders should wear tracking devices so their location can be pinpointed at any time. 13) Only a few sex offenders are dangerous. 14) Most sex offenders are unmarried men. 15) Someone who uses emotional control when committing a sex offense is not as bad as someone who uses 86

PAGE 87

physical control when committing a sex offense. 16) Most sex offenders keep to themselves. 17) A sex offense committed against someone the pe rpetrator knows is less serious than a sex offense committed against a stranger. and 18) Convicted sex offenders should never be released from prison. The CATSO Scale statements are measured on a six item Likert scale, with possible responses listed a s "Strongly Disagree," "Disagree," "Probably Disagree," "Probably Agree," "Agree," and "Strongly Agree." These questions were used with permission from Wesley Church and were used verbatim from the published scale This dependent variable will be used in an OLS Regression, tested against the independent variable of Registry Knowledge. The analytic plan to utilize this dependent variable is explained in Table 3 11 Model 9 as seen at the end of the operationalization portion of this paper. Registry Suppo rt The final community related variable was support for the registry. The survey items either referred to "sexual offender" or "sexual predator" depending on which randomly assigned survey form the respondent received. The last model will examine how sup portive participants are of the sex offender registry in general. These measures will work in combination with the CATSO scale questions to provide adequate attention to the level of support given to the current state of the registry. Registry support is operationalized through the following measures: 1) "Reforms should be made to the sex offender registry." 2) "The sex offender registry is effective in reducing sex offender reoffending." 3) "The sex offender registry makes life difficult for sex offender s living in the community." 4) "Children are safer if the locations of sex offenders are known." 5) "It is justified when individuals retaliate against sex offenders." 87

PAGE 88

6) "Individuals who retaliate against sex offenders should be subject to legal action." 7) "Sex offenders should be released into the community after their prison sentences." 8) "After a certain number of years, a registered sex offender should be able to be removed from the sex offender registry." 9) "It would be too harsh to make sex offend ers wear a special kind of marker, at all times on their person, which identifies them as a sex offender." 10) "Having to register on the sex offender registry constitutes cruel and unusual punishment." 11) "I would support legislative action, which calls for sex offenders to be supervised under electronic monitoring/GPS tracking for the remainder of their lives while living in communities." 12) "I support residency restrictions that prevent sex offenders from living too closely to schools, playgrounds and other areas where children frequently gather." Response options for all twelve questions were measured on a five item Likert scale with responses including, Strongly Disagree, "Disagree," "Unsure," "Agree," "Strongly Agree." These questions were develop ed primarily for this study and were not reproduced from any previous research This dependent variable will be used in an OLS Regression, tested against the independent variable of Registry Knowledge. The analytic plan to utilize this dependent variable is explained in Table 12 Model 3 10 as seen at the end of the operationalization portion of this paper. Demographic and Control Variables Two background variables are given theoretical import in the guiding hypoth eses: parental status and the participa nts' gender. Parental Status was gathered by asking if the participant if he/she was a parent, with a dichotomous "No" or "Yes" response option. Number of School Age Children was collected by asking the participants, "How many children do you currently have enrolled 88

PAGE 89

in either an elementary, middle or high school?" Participants who answered that they were not parents, were assigned "0 children" as their answer. Those who answered that they were indeed parents were given the response options of "1 child, "2 children," "3 children," or "4 or more children." Gender was obtained by asking, "What is your gender?" and was measured dichotomously through the following answer choices, "Male" and "Female." Other Demographic Questions Age of the participants wa s obtained by a sking, "How old are you?" and participants were allowed to chose their appropriate age from a drop down selection of ages ranging from 18 100. Race of the parent participants was obtained by asking, "What is your race?" and was measured c ategorically through the following answer choices, "Native American/Alaskan," "Asian," Native Hawaiian or other Pacific Islander ," "Black/African American," "White," and "Other." Ethnicity was obtained by asking, "Do you identify yourself as Hispanic?" and was measured dichotomously through the following answer choices, "No" and "Yes." Marital status was measured by asking participants, "What is your marital status?" and was measur ed categorically through the following answer choices, "Single," "Married," "Divorced," "Widowed," and "Other (Please Specify)." Education level was measured by asking participants, "What is the highest level of education you have completed?" with respons e options including, "Middle School/Junior High School," "High School/GED Equivalent," "Some College," "Bachelor's Degree," "Master's Degree," "Doctorate," "J.D./Law Degree," or "Other Professional Degree." Current family income was measured by asking th e question, "What is your current family income level?" and was measured categorically through the following 89

PAGE 90

answer choices, "$0 $50,000," "$50,001 $100,000," "$100 001 $150,000," "$150,001 $200,000," and "$200,001 and up." Finally, participants w ere asked about their socio economic standing by asking the question, "What socio economic class do you most identify with?" and this was measured categorically through the following answer choices, "Lower Class," "Middle Class," "Upper Middle Class," and "Upper Class." Population size was collected by asking the participants, "To your best estimate, how many people reside in the area that you live?" with response options including "Less than 50,000," "50,000 99,999," "100,000 249,000," "250,000 999, 999," "1,000,000 or more." Finally, participants were asked about the place that they live. First, they were asked what state they currently reside in. After that information was collected, the states were grouped into the "Northeast," "Midwest," "South and "West" according to the categorization used by the Census Bureau. All of these demographic variables will be used as control variables in the OLS Regression models for the non parent participant group. Initial testing did not produce any significan t relationships between Registry Knowledge and the control variables. These bivariate correlations can be found in the Appendix (Table B 22). Data Analysis Plan The hypotheses suggest the plan of analysis. For all the hypotheses in each category, analysi s begins with a general descriptive statement. Summary descriptive statistics are used to examine, for example, whether only a few or many people access the registries and whether the level of knowledge about registries is high or low. Those general desc riptive statements are then followed by only slightly more exacting predictions that there will be few differences across social categories. Many of those differences can be examined through t tests or by comparing the strength of bivariate 90

PAGE 91

relationships (e.g., whether age relates to knowledge of the registry). The final tier of hypotheses offer s specific expectations about differences (e.g., that parents will be different from non parents in their hostility). To be sure that these predicted differences exist, they are subjected to multivariate analysis. What follows is a series of tables that summarize models for statistical analyses. The descriptive hypotheses do not require specifying a model. The tables, under the analytic plan, will indicate whet her the statistical analysis will be frequency testing (Univariate), bivariate, or multivariate analyses. Most of the tables include multivariate models. The models will identify the variables used and whether they are used as independent or dependent va riables. Tables 3 3 to 3 15 present models that examine the hypotheses relating to personal orientations toward th e control of sexual offenders; Tables 3 5 present models that examine fear of sexual victimization, and Tales 3 7 to 3 13 present models rela ted to community related attitudes. 91

PAGE 92

Table 3 3 Model 1: Across Group Comparison for Registry Website Access Research Question and Hypothesis Independent Variable Dependent Variable Questions for Model Analytic Plan 1) Generally, relatively few people will have ever accessed the registries. 2) Mo re specifically, the extent to which people will have ever access ed the registries will be similar across social categories and groups, and the amount of access will be similar for both the group responding to "sex offender" and that responding to "sex predator" survey items. 2a ) Parents of children will have access ed the registries more often than will non parents. 2b ) Women will have access ed the registries more often than will men. N/A N/A 1) Have you ever looked at your state's sex offender registry website? No (0) Yes (1) 2) Have you ever searched your state's registry website for sex offenders (predators) living in the areas surrounding your home? No (0) Yes (1) 3) How many sex offenders would you estima te are living nearby your home? Drop down list of 0 to 100 offenders Collapse into 10 categories Frequency testing Bivariate analysis T test comparisons (independent sample t test) Due to the nature of the research question, this model is not attempting to predict anything. Rather, it is an examination of how often the sex offender registry is accessed and which group is more frequent in accessing it. 92

PAGE 93

Table 3 4 Model 2: Registry Knowledge Across Participants Research Question and Questions for Model Analytic Plan Registry Knowledge: 1) Generally, the level of knowled ge about registries will be low 2) More specifically, the level of knowledge will be similar across social categories and groups. 2a ) Parents of children will have higher knowledge than non parents. 2b ) Women will have higher knowledge than men. Registry Knowledge: 1) In my state, all sex offenders are classified the same no matter their crime.* (False) 2) Registered sex offenders a re required to live at least 1,0 00 feet from a school zone, park or bus stop. (True) 3) Some sex offenders are required to register for life. (True) 4) Juvenile sex offenders, who are 14 years old at the time of the offense, can be placed on the registry if convicted. (True) 5) All sex offenders are required to be on some sort of electronic monitoring/GPS tracking device at all times*. (False) 6) Sex offenders have very high rates of reoffending*. (False) 7) The Amber Alert System is named after a chil d named Amber, it has nothing to do with the color amber. (True) 8) There are more male sex offenders registered than female sex offenders. (True) 9) Individuals convicted of their very first sexual crime can be classified as sexual predators or can be p laced in a Tier III classification. (True) 10) After serving their prison sentences, sex offenders can be incarcerated indefinitely through the process called Civil Commitment. (True) Answer Choices and Coding: Very True (1) Somewhat True (2) Unsure (3) Somewhat False (4) Very False (5) *Those measures with an asterisk must be reverse coded before analysis. Registry Knowledge: Additive scores for knowledge Create a 10 item count across responses Sets up for a positive relationship between registry knowledge and predictive models Score of 0 indicates no correction responses, whereas a score of 10 indicates 10 correct responses Compare group averages Frequency Testing T test comparisons (unpaired t tests) Due to the nature of the research question, this model is not attempting to predict anything. Rather, it is an examination of the accuracy of participants' knowledge regarding the sex offender registry and the corresponding laws. 93

PAGE 94

Table 3 5: Model 2: Predicting the Stereotypical Sex Offender. Research Question Questions for Model Analytic Plan Stereotypical Sex Offender: 1) Generally, the level of inaccurate stereotyping will be high. 2) More specifically, the level of inaccurate stereotyping will be similar across social categories and groups. 2a ) Those with more knowledge about the registries will have lower levels of inaccurate stereotyping. 2b ) Parents of children will have higher levels of inaccurate stereotyping than will non parents. 2c ) Women will have higher levels of inaccura te stereotyping than will men. Predicting the Stereotypical Sex Offender: 1) Most sex offenders of are of the _____ gender Answer Choices: Male (0) Female (1) 2) and are in the ____ age group Answer Choices: 14 25 years old (1) 26 35 years old (2) 36 45 years old (3) 46 55 years old (4) 56 65 years old (5) 66 76 years old (6) 3) Most sex offenders are of the ____ race Answer Choices: Native American/Alaskan (1) Asian (2) Native Hawaiian or other Pacific Islander (3) Black/African American (4) White (5) Other (6) 4) and are of the ___ ethnicity. Answer Choices: Hispanic (0) Not Hispanic (1) 5) Most sex offenders are ____ to their victims Answer Choices: Stereotypical Sex Offender: Created Count Scores for each of the measures Additive Count score across measures Score of 0 indicates very inaccurate offender profile Score of 7 indicates very accurate offender profile Sets up for a positive relationship between the stereotypical offender profile and predictive models Compare group averages Frequency Testing T test comparisons (unpaired t tests) 94

PAGE 95

Table 3 5 Continued Research Question Questions for Model Analytic Plan 5) Most sex offenders are ____ to their victims Answer Choices: Offender is a stranger (1) Offender is a close friend (2) Offender is a distant relative (3) Offender is an immediate relative (4) 6) and frequently have ____ as their victims. Answer Choices: Pre pubescent female minors (1) Pre pubescent male minors (2) Post pubescent female minors (3) Post pubescent male minors (4) Adult females (5) Adult males (6) 7) ____ is the most frequent form of victimization that sex offenders commit. Physical non consensual sex act (1) Physical consensual sex act with a minor (2) Non physical sex act (3) Due to the nature of the research question, this model is not attempting to predict anything. Rather, it is an examination of the accuracy of participants' knowledge regarding the sex offender registry and the corresponding laws. 95

PAGE 96

Table 3 6 : Model 3: OLS Regression Model Predicting Fear of Victimization Research Question and Hypothesis Independent Variable s Dependent Variable Analytic Plan Personal Victimization 1) Generally, the level of fear of personal victimization will be moderate. 2) More specifically, the level of fear of personal victimization will be similar across social categories and groups. 2a ) N on parent participants will report more fear of personal victimization than will parents. 2b ) Women will report more fear of personal victimization than will men. Victimization of Children 1) Generally, the level of fear of victimization of children will be high 2) More specifically, the level of fear of victimization of children will be similar across social categories and groups. 2a ) Parents will report more fear of victimization of children than will non parents. 2 b ) Women will report more fear of victimization than will men. Registry Knowledge: Ten items as previously discussed in Table 3 4. Measures will be implemented into a 10 item count across responses Stereotypical Sex Offender: Seven items as previously discussed in Table 3 5. Measures will be implemented into a 7 item count variable. Fear of Victimization (Parents and Non Parents) 1) Are you worried that you personally may be a victim of a sexual offense? 2) Are you worried that children in your community may be at risk of becoming victims of a sexual offense? Answer Choices: Definitely Not (1) Probably Not (2) Probably Yes (3) Definitely Yes (4) T test comparisons (unpaired t tests) OLS Regression Scaled IV Single Question DV Run the scaled IV against the two DVs (one for each group) 2 OLS Regressions total 96

PAGE 97

Table 3 7 : Model 4: OLS Regre ssion Model Predicting the Moral Panic Element of Concern Research Question and Hypothesis Independent Variable s Dependent Variable Analytic Plan Concern: 1) Generally, the concern about sexual offenses and offending will be high. 2) More specifically, that concern will be similar across social categories and groups and it will be similar for the group responding to "sex offender" and that gr oup responding to "sex predator" survey items 2a) Parents will more concerned than non parents. 2 b ) Women will be more concerned than men. Registry Knowledge: Ten items as previously discussed in Table 3 4. Measures will be implemented into a 10 item count across responses Stereotypical Sex Offender: Seven items as previously discussed in Table 3 5. Measures will be implemented into a 7 item count variable. Concern (Parent and Non parent Groups) 1) Are you worried about sex offenders (predators) living nearby your home? 2) Are you worried that children in your community may be at risk of becoming t he victim of a sexual offense? 3) Are you worried that you personally may be at risk of becoming the victim of a sexual offense? 4) Are you worried about children in your community being at risk of being approached by a sexual offender ( predator ) ? 5) Ar e you worried that if sex offenders (predators) are living in the community, then more sexual offenses will occur? Answer Choices: Definitely Not (1) Probably Not (2) Probably Yes (3) Definitely Yes (4) OLS Regression Scaled IV Scaled DV Run the scaled IV against the two scaled DVs (one for each group) 2 OLS Regressions total The DV of concern was randomized for participants to either receive questions regarding sexual offenders or sexual predators. The randomization remains constant throughout the sur vey, meaning the participants will only receive questions about offenders or predators. They will never be asked about both groups. 97

PAGE 98

Table 3 8 : Model 5: OLS Regression Model Predicting the Moral Panic Element of Hostility Research Question and Hypothesis Independent Variable s Dependent Variable Analytic Plan 1) Generally, the hostility/anger about sexual offenses and offending will be high. 2) More specifically, that hostility/anger will be similar across social categories and groups and it will be similar for the group responding to "sex offender" and that gr oup responding to "sex predator" survey items 2a ) Parents will be more hostile than non parents. 2 b ) Women will be more hostile than men. Registry Knowledge: Ten items as previous ly discussed in Table 3 4. Measures will be implemented into a 10 item scale across responses Stereotypical Sex Offender: Seven items as previously discussed in Table 3 5. Measures will be implemented into a 7 item count variable. Hostility/Anger 1) Are you angry that sex offenders (predators) are allowed to live in the community? 2) Do you feel any resentment over the fact that some of your neighbors may be sex offenders (predators) ? 3) Do you feel any anger towards the criminal justice system for releasing sex offenders (predators) from jails and prisons? 4) Are you angry that sex offenders (predators) may be working at businesses where you may frequently shop or visit? 5) Are you angry that children in your community might come i nto contact wit h a sex offender (predator) ? Answer Choices: Definitely Not (1) Probably Not (2) Probably Yes (3) Definitely Yes (4) OLS Regression Scaled IV Scaled DV Run the scaled IV against the two scaled DVs (one for each group) 2 OLS Regressions total The DV of hostility was randomized for participants to either receive questions regarding sexual offenders or sexual predators. The randomization remains constant throughout the survey, meaning the participants will only receive questions about offenders or predato rs. They will never be asked about both groups. 98

PAGE 99

Table 3 9 : Model 6: OLS Regression Model Predicting the Moral Panic Element of Consensus Research Question and Hypothesis Independent Variable s Dependent Variable Analytic Plan 1) Generally, the consensus about sexual offenses and offending will be high. 2) More specifically, that consensus will be similar across social categories and groups and it will be similar for the group responding to "sex offender" and that gr oup responding to "sex predator" survey items 2a ) Parents will be more unified than non parents. 2 b ) Women will be more unified than men. Registry Knowledge: Ten items as previously discussed in Table 3 4. Measures will be implemented into a 10 item count across responses Stereotypical Sex Offender: Seven items as previously discussed in Table 3 5. Measures will be implemented into a 7 item count variable. Consensus 1) Do you think that a majority of community members are in agreement about the risk that sex offender s (predators) pose? 2) Do you think that many community members feel that changes must be made in the supervision of sex offenders (predators)? 3) Do you that the community members in general feel threatened by sex offenders (predators) as a group? 4) D o you think that a majority of community members are in agreement that children are at risk of being sexually victimized? 5) Do you think that many community members feel that sex offenders (predators) are too dangerous to be living in the community? Ans wer Choices: Definitely Not (1) Probably Not (2) Probably Yes (3) Definitely Yes (4) OLS Regression Scaled IV Scaled DV Run the scaled IV against the two scaled DVs (one for each group) 2 OLS Regressions total The DV of consensus was randomized for participants to either receive questions regarding sexual offenders or sexual predators. The randomization remains constant throughout the survey, meaning the participants will only receive questions about offenders or predators. They will never be asked about both groups. 99

PAGE 100

Table 3 10 : Model 7: OLS Regression Model Predicting the Moral Panic Element of Disproportionality Research Question and Hypothesis Independent Variable s Dependent Variable Analytic Plan 1) Generally, the reaction about sexual offenses and offending will be disproportionate. 2) More specifically, the disproportionate responses will be similar across social categories and groups and it will be similar for the group responding to "sex offender" and that gr oup responding to "sex predator" survey items 2a ) Parents will react more disproportionately than will non parents. 2 b ) Women will act more disproportionately than will men. Registry Knowledge: Ten items as previously discussed in Table 3 4. Measures will be implemented into a 10 item count across responses Stereotypical Sex Offender: Seven items as previously discussed in Table 3 5. Measures will be implemented into a 7 item count variable. Disproportionality: 1) Do you feel that the current state of the sex offend er registry is too harsh?* 2) Do you feel that the sex offender registry laws should be stricter? 3) Do you feel that keeping sex offenders (predators) on electronic monitoring/GPS tracking for more than 5 years without a break is too severe a punishment ?* 4) Do you feel that sex offenders (predators) should report to law enforcement more than the required two times per year? 5) Do you feel that the media overreacts in their reporting of sex offenses when they occur in a community?* Answer Choices: Definitely Not (1) Probably Not (2) Probably Yes (3) Definitely Yes (4) *Those measures with an asterisk must be reverse coded before analysis. OLS Regression Scaled IV Scaled DV Run the scaled IV against the two scaled DVs (one for each group) 2 OLS Re gressions total The DV of disproportionality was randomized for participants to either receive questions regarding sexual offenders or sexual predators. The randomization remains constant throughout the survey, meaning the participants will only receive questions about offenders or predators. They will never be asked about both groups. 100

PAGE 101

Table 3 11 : Model 8: OLS Regression Model Predicting the Moral Panic Element of Volatility Research Question and Hypothesis Independent Variable s Dependent Variable Analytic Plan To be explored, no a priori hypotheses advance Registry Knowledge: Ten items as previously discussed in Table 3 4. Measures will be implemented into a 10 item count across responses Stereotypical Sex Offender: Seven items as previously discussed in Table 3 5. Measures will be implemented into a 7 item count variable. Volatility: 1) Do you feel that law enforcement reacts quickly enough when a sexual offense takes place? 2) Do you feel that legislators work fast enough to get necessary registry laws passed to further keep track of sex offenders (predators)? 3) Do you feel that the media reports sex offenses cases too quickly before all of the facts are gathered? 4) Do you feel that the quick response of the media makes communities saf er because people are made aware of the sex offense? 5) Do you feel that police are too slow to catch sex offenders when sexual offenses take place? Answer Choices: Definitely Not (1) Probably Not (2) Probably Yes (3) Definitely Yes (4) OLS Regression S caled IV Scaled DV Run the scaled IV against the two scaled DVs (2 OLS Regressions total) The DV of volatility was randomized for participants to either receive questions regarding sexual offenders or sexual predators. The randomization remains constant throughout the survey, meaning the participants will only receive questions about offenders or predators. They will never be asked about both groups. 101

PAGE 102

Table 3 12 : Model 9 : OLS Regression Model Predicting the Community Attitudes Towards Sex Offenders (CATSO) Scale Research Question and Hypothesis Independent Variable Dependent Variable Analytic Plan 1) Generally, citizens will have highly negative attitudes for the overall scale and on each of the four sub dimensions (Social Isolation, Capacity for Change, Severity/Dangerousness, and Deviancy). 2) More specifically, these attitudes will be similar across social categories and groups 2a ) Those with registry knowledge will hold more negative attitudes on the CATSO scale and sub dimensions than will t hose with less registry knowledge. 2 b ) Parents will hold more negative attitudes on the CATSO scale and sub dimensions than non parents. 2c ) Women will hold more negative attitudes on the CATSO scale and sub dimensions than will men. Registry Knowledge: Ten items as previously discussed in Table 3 4. Measures will be implemented into a 10 item count across responses Stereotypical Sex Offender: Seven items as previously discussed in Table 3 5. Measures will be implemented into a 7 item count variable CATSO Scale 1) With support and therapy, someone who committed a sexual offense can learn to change their behavior.* 2) People who commit sex offenses should lose their civil rights (e.g. voting and privacy). 3) People who commit sex offenses want to have sex more often then the average person. 4) Male sex offenders should be punished more severely then female sex offenders.* 5) A lot of sex offenders use their victims to create pornography. 6) Sex offenders prefer to stay home alone rather than to be around lots of people. 7) Most sex offenders do not have close friends. 8) Sex offenders have difficulty making friends even if they try real hard. 9) The prison sentences sex offenders receive are much too long when compared to the sentence lengths for other crimes.* 10) Sex offenders have high rates of sexual activity. 11) Trying to rehabilitate sex offenders is a waste of time. 12) Sex offenders should wear tracking devices so that their location can be pinpointed at any time. OLS Regression Scaled IV Scaled DV into four constructs Total index of four scales Run the scaled IV against the two scaled DVs (5 OLS Regressions total) 102

PAGE 103

Table 3 12. Continued Research Question and Hypothesis Independent Variable Dependent Variable Analytic Plan 13) Only a few sex offenders are dangerous.* 14) Most sex offenders are unmarried men. 15) Someone who uses emotional control when committing a sex offense is not as bad as someone who uses physical control when committing a sex offense.* 16) Most sex o ffenders keep to themselves. 17) A sex offense committed against someone the perpetrator knows is less serious than a sex offense committed against a stranger.* 18) Convicted sex offenders should never be released from prison. Answer Choices: Strongly Disagree (1) Disagree (2) Probably Disagree (3) Probably Agree (4) Agree (5) Strongly Agree (6) *Those measures with an asterisk must be reverse coded before analysis. 103

PAGE 104

Table 3 13 : Model 10 : OLS Regression Model Predicting Registry Support Research Question and Hypothesis Independent Variable Dependent Variable Analytic Plan 1) Generally, citizens will show high support for the registry. 2) More specifically, the support will be similar across social categories and groups and it will be similar for the group responding to "sex offender" and that group responding to "sex predator" survey items 2a ) Parents will be more supportive of the registry than will non parents. 2 b ) Women will be more supportive of the registry than will men. Registry Knowledge: Ten items as previously discussed in Table 3 4. Measures will be implemented into a 10 item count across responses Stereotypical Sex Offender: Seven items as previously discussed in Table 3 5. Measures will be implemented into a 7 item count variable. Registry Support 1) Reforms should be made to the sex offender registry 2) The sex offender registry is effective in reducing sex offender reoffending. 3) The sex offender registry makes life very difficult for sex offenders living in the community. 4) Children are safer if the locations of sex offenders are known. 5) It is justified when individuals retaliate against sex offenders. 6) Individuals who retaliate against sex offenders should be subject to legal action. 7) Sex offenders should be released into the community after their prison sentences. 8) After a certain number of years, a registered sex offender should be able to be removed from the sex offender registry. 9) It would be too harsh to make sex offenders wear a special kind of marker, at all times on their person, which identifies them as a sex offender. 10) Having to register on the sex offender registry constitutes cruel and unusual punishment. 11) I would support legislative action which calls for sex offen ders to be supervised under electronic monitoring/GPS tracking for OLS Regression IV Scaled DV Run the IV against the two scaled DVs (2 OLS Regressions total) The DV of registry support was randomized for participants to either receive questions regard ing sexual offenders or sexual predators. The randomization remains constant throughout the survey, meaning the participants will only receive questions about offenders or predators. They will never be asked about both groups. 104

PAGE 105

Table 3 13. Continued Research Question and Hypothesis Independent Variable Dependent Variable Analytic Plan the remainder of their lives while living in communities. 12) I support residency restrictions that prevent all sex offenders from living too closely to schools, playgrounds and other areas where children frequently gather. Answer Choices: Strongly Disagree (1) Disagree (2) Unsure (3) Agree (4) Strongly Agree (5) 105

PAGE 106

CHAPTER 4 RESULTS Following the previously discussed data analysis plans, this section of the paper displays the results of those plans. This section of the dissertation will begin with participant demographics and then segues into the analysis of Models 1 10. Participant Demographics As previously discussed, 877 surveys were completed via the MTurk survey participation system. In Table 4 1, the participant demographics show nearly an even distribution of male (n = 465, 53.0%) and female (n = 412, 47.0%) participants. Participant ages are a bit more dispersed with the majority of participant between the ages of 18 39. The majority of the participants identified as being white (n = 521, 59.4%) or Asian (n = 248, 28.3%). In regards to ethnicity, most participants did not identify as be ing Hispanic (n = 781, 90.0%). Table 4 1. Univariate Analysis for Participant Demographics: Gender, Age, Race and Ethnicity Measure Code Frequency Percent Gender 0 = Male 1 = Female n = 465 n = 412 53.0% 47.0% Age 1 = 18 24 2 = 25 29 3 = 30 34 4 = 35 39 5 = 40 44 6 = 45 49 7 = 50 54 8 = 55 59 9 = 60 64 10 = 65 and older n = 205 n = 219 n = 142 n = 84 n = 64 n = 34 n = 47 n = 42 n = 25 n = 15 23.4% 25.0% 16.2% 9.6% 7.3% 3.9% 5.4% 4.8% 2.9% 1.7% Race 1 = Native American/Alaskan 2 = Asian 3 = Native Hawaiian or other Pacific Islander 4 = Black/African American 5 = White 6 = Other n = 5 n = 248 n = 3 n = 62 n = 521 n = 38 0.6% 28.3% 0.3% 7.1% 59.4% 4.3% Ethnicity 0 = Not Hispanic 1 = Hispanic n = 781 n = 96 89.1 % 10.9% 106

PAGE 107

Table 4 2. Univariate Analysis for Participant Demographics: Parental Status, Information Regarding Children, and Grandparental Status Measure Code Frequency Percentage Parental Status 0 = Not a parent 1 = P arent n = 491 n = 386 56.0% 44.0% Number of children 0 = No children 1 = 1 child 2 = 2 children 3 = 3 children 4 = 4 or more children n = 386 n = 219 n = 154 n = 82 n = 36 44.0% 25.0% 17.6% 9.4% 4.1% (Total 100%) Number of children enrolled in school 0 = No children 1 = 1 child 2 = 2 children 3 = 3 children 4 = 4 or more children n = 611 n = 153 n = 78 n = 29 n = 6 69.7% 17.4% 8.9% 3.3% 0.7% (Total 100%) Age of First Child 1 = 0 4 years old 2 = 5 9 years old 3 = 10 14 years old 4 = 15 18 years old 5 = 19 and older Missing (were not displayed question) n = 174 n = 105 n = 61 n = 34 n = 116 n = 386 19.9% 12.0% 7.0% 3.9% 13.2% (Total 56%) 44.0% Age of Second Child 1 = 0 4 years old 2 = 5 9 years old 3 = 10 14 years old 4 = 15 18 years old 5 = 19 and older Missing (were not displayed question) n = 34 n = 54 n = 40 n = 20 n = 71 n = 658 3.9% 6.1% 4.6% 2.3% 8.1% (Total 25%) 75.0% Age of Third Child 1 = 0 4 years old 2 = 5 9 years old 3 = 10 14 years old 4 = 15 18 years old 5 = 19 and older Missing (were not displayed question) n = 22 n = 29 n = 37 n = 24 n = 42 n = 723 2.5% 3.3% 4.2% 2.7% 4.8% (Total 17.5%) 82.5% Age of Fourth Child 1 = 0 4 years old 2 = 5 9 years old 3 = 10 14 years old 4 = 15 18 years old 5 = 19 and older Missing (were not displayed question) n = 3 n = 2 n = 8 n = 6 n = 17 n = 841 0.3% 0.2% 0.9% 0.8% 1.9% (Total 4.1%) 95.9% Number of Male Children 0 = No children 1 = 1 child 2 = 2 children 3 = 3 children 4 = 4 or more children Missing (were not displayed question) n = 230 n = 192 n = 59 n = 10 n = 0 (Total 261) n = 386 26.2% 21.9% 6.8% 1.1% 0.0% (Total 54.0%) 44.0% 107

PAGE 108

Table 4 2. Continued. Measure Code Frequency Percentage Number of Female Children 0 = No children 1 = 1 child 2 = 2 children 3 = 3 children 4 = 4 or more children Missing (were not displayed question) n = 277 n = 172 n = 37 n = 5 n = 0 (Total 214) n = 386 31.6% 19.6% 4.2% 0.6% 0.0% (Total 54%) 44.0% Intend to Have More Children 0 = No 1 = Yes n = 720 n = 157 82.1% 17.9% Grandparental Status 0 = No 1 = Yes n = 813 n = 64 92.7% 7.3% Participants were also asked if they had any children. Like gender, parental status had a close distribution between parents (n = 491, 56.0%) and non parents (n = 386, 44.0%). Of those individuals who had children, the majority of participants only have one (n = 219, 25.0%) or two children (n = 154, 17.6%). A large number of participants did not have a child who was enrolled in school (n = 611, 69.7%), and 17.4% (n = 153) of participants had at least one child enrolled in school. Participants were asked about the ages of their children and for the most part, those children were younger than 19 years of age which coincides with the enrollment in school variable. Overall, parents reported that they have more male children (n = 261, 29.8%, collective tot al for options of 1, 2, 3, and 4 or more children) than female children (n = 217, 24.4%, collective total for options of 1, 2, 3 and 4 or more children). Participants overwhelming reported that they did not intend to have any more children (n =720, 82.1%) and very few individuals reported to be grandparents (n = 64, 7.3%) The remaining social demographic questions asked participants about their education level, marital status, income level and socio economic status. Most participants completed at least so me collect (n = 273, 31.1%) or have a Bachelor's Degree (n = 368, 42.0%). Participants were mainly single (n = 386, 44.0%) or married 108

PAGE 109

(n = 416, 47.4%) and earned less than $50 thousand a year (n = 512, 58.4%). Based on that income level, the majority of participants identified themselves as being middle class (n = 593, 67.6%). Table 4 3. Univariate Analysis for Participant Demographics: Education, Marital Status, Income Level, Socio Economic Status Measure Code Frequency Percentage Highest Level of Education 1 = Middle School/Junior High School 2 = High School/GED Equivalent 3 = Some College 4 = Bachelor's Degree 5 = Master's Degree 6 = Doctorate 7 = J.D./Law Degree 8 = Other Professional Degree n = 5 n = 79 n = 273 n = 368 n = 130 n = 8 n = 9 n = 5 0.6% 9.0% 31.1% 42.0% 14.8% 0.9% 1.0% 0.6% Marital Status 1 = Legally Single 2 = Married 3 = Divorced 4 = Widowed n = 386 n = 416 n = 64 n = 11 44.0% 47.4% 7.3% 1.2% Family Income Level 1 = $0 $50,000 2 = $50,001 $100,000 3 = $100,001 $150,000 4 = $150,001 $200,000 5 = $200,001 and higher n = 512 n = 275 n = 66 n = 19 n = 5 58.4% 31.4% 7.5% 2.2% 0.6% Socio Economic Class 1 = Lower Class 2 = Middle Class 3 = Upper Middle Class 4 = Upper Class n = 171 n = 593 n = 110 n = 3 19.5% 67.6% 12.5% 0.3% Since the goal of this study was to collect a nation wide sample, participants were asked about their geographic demographics (Table 4 4). At least one participant was collected from every state and the District of Columbia and the most frequent states of residence were California (n = 83, 9.5%), Florida (n = 61, 7.0%), Indiana (n = 79, 9.0%), New York (n = 64, 7.3%) and Texas (n = 60, 6.8%). There is also a relatively even distribution of participants living in rural, urban and suburban areas, with urban environments being slightly more frequent (n = 345, 39.3%). Population wise, participants reported less often that they lived in an area where the population was greater than 250,000 people (n = 184, 21.0%, collective total for higher population options) 109

PAGE 110

Table 4 4. Univariate Analysis for Participant Demographics: State of Residence, Type of Residence, Population Size, Geographic Region Measure Code Frequency Percentage State of Residence 1 = Alabama 2 = Alaska 3 = Arizona 4 = Arkansas 5 = California 6 = Colorado 7 = Connecticut 8 = Delaware 9 = District of Columbia 10 = Florida 11= Georgia 12 = Hawaii 13 = Idaho 14 = Illinois 15 = Indiana 16 = Iowa 17 = Kansas 18 = Kentucky 19 = Louisiana 20 = Maine 21 = Maryland 22 = Massachusetts 23 = Michigan 24 = Minnesota 25 = Mississippi 26 = Missouri 27 = Montana 28 = Nebraska 29 = Nevada 30 = New Hampshire 31 = New Jersey 32 = New Mexico 33 = New York 34 = North Carolina 35 = North Dakota 36 = Ohio 37 = Oklahoma 38 = Oregon 39 = Pennsylvania 40 = Rhode Island 41 = South Carolina 42 = South Dakota 43 = Tennessee 44 = Texas 45 = Utah 46 = Vermont 47 = Virginia 48 = Washington 49 = West Virginia 50 = Wisconsin 51 = Wyoming n = 12 n = 1 n = 24 n = 5 n = 83 n = 15 n = 10 n = 5 n = 3 n = 61 n = 30 n = 3 n = 5 n = 31 n = 79 n = 7 n = 5 n = 16 n = 9 n = 1 n = 16 n = 16 n = 27 n = 11 n = 5 n = 10 n = 2 n = 5 n = 11 n = 2 n = 14 n = 5 n = 64 n = 30 n = 2 n = 27 n = 7 n = 14 n = 33 n = 2 n = 15 n = 2 n = 16 n = 60 n = 10 n = 2 n = 16 n = 20 n = 10 n = 17 n = 1 1.4% 0.1% 2.7% 0.6% 9.5% 1.7% 1.1% 0.6% 0.3% 7.0% 3.4% 0.3% 0.6% 3.5% 9.0% 0.8% 0.6% 1.8% 1.0% 0.1% 1.8% 1.8% 3.1% 1.3% 0.6% 1.1% 0.2% 0.6% 1.3% 0.2% 1.6% 0.6% 7.3% 3.4% 0.2% 3.1% 0.8% 1.6% 3.8% 0.2% 1.7% 0.2% 1.8% 6.8% 1.1% 0.2% 1.8% 2.3% 1.1% 1.9% 0.1% 110

PAGE 111

Table 4 4. Continued. Measure Code Frequency Percentage Type of Residential Area 1 = Rural 2 = Urban 3 = Suburban n = 200 n = 345 n = 332 22.8% 39.3% 37.8% Population Size 1 = Less than 50,000 2 = 50,000 99,999 3 = 100,000 249,000 4 = 250,000 999,999 5 = 1,000,000 or more n = 238 n = 245 n = 210 n = 100 n = 84 27.1% 27.9% 23.9% 11.4% 9.6% Geographic Region 1 = Northeast 2 = Midwest 3 = South 4 = West n = 144 n = 223 n = 316 n = 194 16.4% 25.4% 36.0% 22.1% After breaking down the state residences further into geographic regions, the South is represented at a slightly higher frequency (n = 316, 36.0%) than the Northeast (n = 144, 16.4 %), Midwest (n = 223, 25.4%), and West (n = 194, 22.1%) regions of the country. Other researchers have estimated that the majority of MTurk workers are female (70%), are between the age of 21 35 (54%), are frequently childless (55%), have income levels th at are less than $60K (65%) (Ipeirotis, 2013), have at least some college experience (21.0%) or have a Bachelor's degree (42.0%) (Ross, Zaldivar, Irani and Tomlinson, 2010). Based on these findings from other researchers, the participant demographics for this study fit within the parameters of prior research. Initial bivariate analysis testing did not find any significant relationships between these control variables and the main independent variable of Registry Knowledge. A full report of those bivariat e analyses can be found in the Appendix in Table B 22. Hypothesis Testing and Results 6 This portion of the results section discusses the ten predicted models and the associated hypotheses. 6 All of the hypothesis testing will follow the proposed order of the hypotheses as discussed in Chapter 3. 111

PAGE 112

Hypothesis Testing Concerning Personal Orientations Toward the Control of Sex Offenders Models Model 1: Across Group Comparison for Sex Offender Registry Website Access In Model 1, the following sets of analyses test two specific hypotheses, which have been derived from more general predictions. First, it is predicte d that few people will have ever accessed the registries (Hypothesis 1a) Second, the extent to which people have ever access ed the registries will be similar across most social categories and groups (1b) Third it is predicted that parents of children will have access ed the registries more often than will non parents (1c) Fourth it is predicted that women will have access ed the registries more often than will men (1d) Table 4 5 shows the u nivariate statist ics for which participants have accessed their states' registry websites. All 877 participants were presented with the questions of whether or not they have ever accessed their state's sex offender registry. A little over half of the sample re ported any kind of access. This does not support the first hypothesis which predicted that only a few people will have ever accessed the registry (1a) Table 4 5. Univariate Statistics for Registry Website Access Measure Code Frequency Percent Have you ever looked at your state's sex offender registry website? 0 = No 1 = Yes n = 457 n = 420 52.1% 47.9% The next step was to conduct bivariate correlations on registry website access and all of the control variables used in this model. Table B 23 (Appendix) shows the results for across group comparison of accessing the states' registry websites and the control variables Several of the control variables show a significant correlation between itself and accessing the registry. N umber of school age children (r = .082) education level (r = .072) marital status (r = .084) socio economic status (r = .074) 112

PAGE 113

geographic region (r = .072) and population size (r = .073) (all significant at .05 alpha level) all show a significant correlation with registry access. This does not support the second part of the hypothesis, which suggested that registry access would similar across social categories and groups (1b). Parent st atus and gender do not have significant correlations with registry access, which does not provide support for either hypothesis (1c and 1d). After examining the u nivariate and bivariate analysis for registry access two independent sample t tests were cond ucted to see if there was any significant difference between parents and non parents and male and female participants respectively in their history of registry access Shown in Table 4 6 the results of the parent t test suggest that there is a significa nt d ifference in registry access between parents (M=.51 SD=.505) and non parents (M=.44 SD= .497 ); t(450)= 2.174, p=.012** The results of this t test suggest that parents have more experience accessing the sex offender registry than do non parents, whic h is supportive of the parent hypothesis (1c) which suggests that exact relationship. The second independent sample t test focusing on gender, suggest that suggest that there is a significant difference in registry access between male (M=.45 SD= .498 ) a nd female participants (M=.51, SD=.505); t(423)= 1.889 p=. 03 9 This t test suggests that females have more experience accessing the registry than do men, which is also supportive of the gender hypothesis (1d) which suggests that relationship as well. Registry access was meant to be a predictor variable for the multivariate models, as it was hypothesized that increase experience in accessing the registry would lead to an increased amount of registry knowledge. However, initial bivariate analysis betwee n those two variables did not show a 113

PAGE 114

significant relationship, so registry access was left out of the model. Furthermore, registry access is more of a behavioral measure and would not be the best fit in trying to predict attitudinal models. Table 4 6 Independent Sample T Test for Parental Status and Gender for Registry Access Registry Access Mean SD N df T Sig. Parents .51 .505 479 450 2.174 012** Non Parents .44 .497 398 450 2.174 Registry Access Male Female .45 .51 .498 .505 465 412 450 450 1.889 1.889 03 9 *** p<.001 **p < .01 *p < .05 Model 2: Registry Knowledge Across Participants and Predicting the Stereotypical Sex Offender For Model 2, it is hypothesized that knowledge about registries will be low (2a) Furthermore, the level of knowledge will be similar for participants across social categories and groups (2b) More specifically, it is hypothesized that parents of children will have higher levels of knowledge than non parents (2c) and that women will have higher levels of knowledge than men (2d) Table 4 7 shows the u nivariate analysis for the ten measures that were used to test the participants' registry knowledge. The univariate a nalysis show s for the most part, participants are largely unsure about many of the se testing measures. There are only a few measures where a definitive response was made by the majority of the participants which expressed a strong answer to the question. For example, for the statement, "The Amber Alert System is named after a child named Amber, it has nothing to do with the color amber," participants largely answered that this statement was very true (n = 436, 49.7%) or somewhat true (n = 135, 15.4%). 114

PAGE 115

Table 4 7 Univariate Statistics for Registry Knowledge Across Participants Meas ure Code Frequency Percentages In my state, all sex offenders are classified the same no matter their crime.* (False) 1 = Very False 2 = Somewhat False 3 = Unsure 4 = Somewhat True 5 = Very True n = 70 n = 135 n = 370 n = 215 n = 87 8.0% 15.4% 42.2% 24.5% 9.9% Registered sex offenders are required to live at least 1,000 feet from a school zone, park or bus stop. (True) 1 = Very True 2 = Somewhat True 3 = Unsure 4 = Somewhat False 5 = Very False n = 196 n = 259 n = 330 n = 60 n = 32 22.3% 29.5% 37.6% 6.8% 3.6% Some sex offenders are required to register for life. (True) 1 = Very True 2 = Somewhat True 3 = Unsure 4 = Somewhat False 5 = Very False n = 369 n = 222 n = 218 n = 48 n = 20 42.1% 25.3% 24.9% 5.5% 2.3% Juvenile offenders, who are at least 14 years old at the time of the offense, can be placed on the registry if convicted. (True) 1 = Very True 2 = Somewhat True 3 = Unsure 4 = Somewhat False 5 = Very False n = 120 n = 189 n = 473 n = 73 n = 22 13.7% 21.6% 53.9% 8.3% 2.5% All sex offenders are r equired to be on some sort of electronic monitoring/GPS tracking device at all times.* (False) 1 = Very False 2 = Somewhat False 3 = Unsure 4 = Somewhat True 5 = Very True n = 238 n = 221 n = 261 n = 93 n = 64 27.1% 25.2% 29.8% 10.6% 7.3% Sex offenders have very high rates of reoffending.* (False) 1 = Very False 2 = Somewhat False 3 = Unsure 4 = Somewhat True 5 = Very True n = 41 n = 68 n = 230 n = 325 n = 213 4.7% 7.8% 26.2% 37.1% 24.3% The Amber Alert system is named after a child named Amber; it has nothing to do with the color amber. (True) 1 = Very True 2 = Somewhat True 3 = Unsure 4 = Somewhat False 5 = Very False n = 436 n = 135 n = 229 n = 43 n = 34 49.7% 15.4% 26.1% 4.9% 3.9% There are more male sex offenders registered than female sex offenders. (True) 1 = Very True 2 = Somewhat True 3 = Unsure 4 = Somewhat False 5 = Very False n = 416 n = 236 n = 164 n = 49 n = 12 47.4% 26.9% 18.7% 5.6% 1.4% Individuals convicted of their very first sexual crime can be classified as sexual predators or can be placed in a Tier III classification. (True) 1 = Very True 2 = Somewhat True 3 = Unsure 4 = Somewhat False 5 = Very False n = 140 n = 223 n = 451 n = 47 n = 16 16.0% 25.4% 51.4% 5.4% 1.8% After serving their prison sentences, sex offenders can be incarcerated indefinitely though a process called Civil Commitment. (True) 1 = Very True 2 = Somewhat True 3 = Unsure 4 = Somewhat False 5 = Very False n = 87 n = 143 n = 524 n = 91 n = 32 9.9% 16.3% 59.7% 10.4% 3.6% Responses are reverse coded 115

PAGE 116

This statement is true and is reflective of accurate registry knowledge among participants. However for other measures, such as "Sex offenders have very high rates of reoffending," the majority of the respondents answered that this was either somewhat true (n = 325, 37.1%) or very true (n = 213, 24.3%). This is an example of a false statement, but participants believe the opposite to be true and shows an inaccurate knowledge base among participants. There are mixed results in regards to the overall level of k nowledge among participants something to be further examined in the remaining analyses for this model. Each of the measur es were dichotomized after the u nivariate analyses were conducted. For the true statement measures, the responses Very True and Some what True were collapsed and coded as Correct. The remaining three categories (Unsure, Somewhat False and Very False) were collapsed as Incorrect. The reverse action was conducted if the measure was a false statement. Once dichotomized, the measures wer e then transformed into a count variable. T able 4 8 Univariate Analysis for Participants' Count Scores for Registry Knowledge Additive Index of Ten Measures Frequency Percent Measure Score of: 0 correct n = 37 4.2% 1 correct n = 41 4.7% 2 correct n = 77 8.8% 3 correct n = 107 12.2% 4 correct n = 169 19.3% 5 correct n = 165 18.8% 6 correct n = 139 15.8% 7 correct n = 84 9.6% 8 correct n = 35 4.0% 9 correct n = 5 0.6% 10 correct n = 18 2.1% With this count variable, a range of 0 10 was produced. In order to have truly accurate knowledge of the registry, the ideal additive response across all ten answers would be a score or a response of 10 indicating that the participant got all ten 116

PAGE 117

measur es correct. A score of 0 would indicate that the participant got none of the measures correct. This sets up a positive relationship between registry knowledge and all of the predicted moral panic elements. Table 4 8 shows the Univariate analysis for the participants' additive scores for the Registry Knowledge variable. Figure 4 1: Additive Responses for Registry Knowledge Measures Figure 4 1 shows a bell curve of responses across participants. As a reminder, a score of 0 indicates less accurate knowle dge, whereas a score of 10 indicates more accurate knowledge. The left end of the bell curve accounts for those participants with low accurate registry knowledge and the right end of the curve accounts for those with highly accurate registry knowledge. 1 69 participants (19.3%) are at the peak of the curve, scoring a 4 across the 10 measures. Only 18 participants (2.1%) exhibited truly accurate knowledge, and the results of this data suggest that there is some inaccurate knowledge being perceived as truth by the participants. Since the majority of the 117

PAGE 118

participants got more than half of the answers wrong, these results support the first hypothesis of the model, which suggests that overall the level of knowledge will be low (2a). After the u nivariate analyses were complete, bivariate correlations were run on registry knowledge and the control variables. No significant correlations were produced from that analysis. The full correlation table can be found in the Appendix (Table B 22). The la ck of significant correlations fully provide support for the second hypothesis regarding registry knowledge, which states that the level of knowledge will be similar across all social groups and categories. The next step in this analysis rests in conducti ng an independent sample t test of additive registry knowledge among the parent and non parent group and among genders Shown in Table 4 9 the results of the t test suggest that there is a weak significant dif ference in registry knowledge between paren ts (M=4.53, SD=2.04) and non parents (M=4.46, SD=2.23); t(875)= .473, p=.049* This infers that parents have higher levels of knowledge compared to the non parent group The results of the independent sample t test support the parent hypothesis which pr edicted that parents would have higher levels of knowledge than non parent s (2c) A second t test was run to compare the accurate knowledge between male and female participants. The results of that t test show a weak significant difference in registry kn owledge between male participants (M= 4.35, SD=2.18 ) and female participants (M= 4.66 SD =2.06); t(875)= 2.131, p=.033* These results show that females have higher levels knowledge based on their overall higher mean responses, providing the groundwork for the positive 118

PAGE 119

relationship between accurate knowledge and all of the predicted models (2d) This gender difference may prove important for the rest of the models overall. Table 4 9 Independent Sample T Test for Parental Status or Gender and Additive Registry Knowledge Additive Knowledge Mean SD N df T Sig. Parents 4.53 2.04 479 875 .473 .049* Non Parents 4.46 2.23 398 875 .473 Additive Knowledge df T Sig. Mean SD N Male 4.35 2.18 4 64 875 2.131 033* Female 4.66 2.06 412 875 2.131 *** p<.001 **p < .01 *p < .05 Predicting the Stereotypical Sex Offender The next step focusing on orientations regarding control is the "popular" knowledge stereotypical sex offender as perceived by the participants Three additional hypotheses have been derived concerning the participants' assumptions about the stereotypical sex offender. Generally, the level of inaccurate stereotyping will be high (3a) For this model, it is hypothesized that the level of inaccurate stereotyping will be similar across social categories and groups (3b) Second, p arents of children will have higher levels of inaccurate ste reotyping than will non parents (3c) Finally, it is hypothesized that women will hav e higher levels of inaccurate stereotyping than will men (3d) This set of questions did not implement the random assignment of "sexual offender" and "sexual predator" questions to the participants. In order to test these hypotheses, frequency analyses, independent sample t tests and OLS Regressions will be completed. Table 4 10 shows the frequency analysis for the stereotypical sex offender measures. Seven measures were used to examine the offender profile that participants believed to be an accurate p ortrayal of who the stereotypical sex offender really is. 119

PAGE 120

Table 4 10 Univariate Analysis for the Stereotypical Sex Offender Measures. Measure (n = 877) Code Frequency Percent Most sex offenders are of the ______ gender 0 = Male 1 = Female n = 834 n = 43 95.1% 4.9% and are in the _____ age group. 1 = 14 25 years old 2 = 2 6 35 years old 3 = 36 45 years old 4 = 46 55 years old 5 = 56 65 years old 6 = 66 75 years old n = 54 n = 406 n = 303 n = 101 n = 11 n = 2 6.2% 46.3% 34.5% 11.5% 1.3% 0.2% Most sex offenders are of the _____ race 1 = Native American/Alaskan 2 = Asian 3 = Native Hawaiian or Other Pacific Islander 4 = Black/African American 5 = White 6 = Other n = 45 n = 76 n = 21 n = 81 n = 627 n = 27 5.1% 8.7% 2.4% 9.2% 71.5% 3.1% and are of the _____ ethnicity. 0 = Hispanic 1 = Not Hispanic n = 163 n = 714 18.6% 81.4% Most sex offenders are ____ to their victims 1 = Offender is a stranger 2 = Offender is a close friend 3 = Offender is a distant relative 4 = Offender is an immediate relative n = 285 n = 257 n = 245 n = 90 32.5% 29.3% 27.9% 10.3% and frequently have ____ as their victims. 1 = Pre pubescent female minors 2 = Pre pubescent male minors 3 = Post pubescent female minors 4 = Post pubescent male minors 5 = Adult females 6 = Adult males n = 361 n = 163 n = 203 n = 19 n = 95 n = 36 41.2% 18.6% 23.1% 2.2% 10.8% 4.1% ____ is the most frequent form of victimization that sex offenders commit. 1 = Physical non consensual sex act 2 = Physical consensus sex act with a minor 3 = Non physical sex act n = 560 n = 228 n = 89 63.9% 26.0% 10.1% Prior literature has suggested that the most frequent registered sex offender is male (97.7% of the total registered sex offenders nationwide), white (67.0%), has a mean age of 44.3 years of age, had a female victim (87.0%) betwe en the ages of 11 14 120

PAGE 121

( 37 .0%), and committed an offense against a child (55.0%) (Ackerman, Harris, Levenson & Zgoba, 2011). Taking this national profile into account, the stereotypical sex offender measures included gender, age, race, and ethnicity of the offender, victim/offender relationship, most frequent victim, and most frequent type of sex offense committed. The stereotypical variables were measured using dichotomous or categorical response op tions. In Table 4 10, t he results show that for the most part, participants are providing a somewhat consistent profile to what was reported in the Ackerman et al. (2011) study. Participants identify the stereotypical offender to be male (95.1%), under t he age of 45 (87.7% cumulative total of the 14 25, 26 35, 36 45 year age groups), white (71.5%), and Not Hispanic (81.4%). Furthermore, participants identified that the offender and victim knew each other prior to the sex offense (67.5% cumulative tot al of the offender being a close friend, distant relative or immediate relative). The most frequent victim was identified as a pre pubescent female minor (41.2%) and that the most frequent sex offense is a physical, non consensual sex act (63.9%). These frequency analyses are the first indication that participants are identifying an accurate offender profile comparative to prior literature. Since the stereotypical sex offender variable will be used as a predictor in the OLS Regressions, it was important to create a count variable across the measures to see how accurate the participants were in identifying the stereotypical sex offender. Each of the seven measures were dichotomized into "correct" and "incorrect" response options, then a new measure was created by indexing the seven variables. Participant scores could range from 0 (no correct responses were identified) to 7 (all seven correct 121

PAGE 122

responses were identified). Table 4 11 shows the u nivariate analysis of the index derived from the seven stereotypical sex offender variables. Less than 10% of participants were able to identify the stereotypical sex offender with full accuracy. Table 4 11 Univariate Analysis for the Variable Index for the Stereotypical Sex Offender Measures. Additive Index of Seven Frequency Percent Measure Score of: 0 correct n = 3 .3% 1 correct n = 24 2.7% 2 correct n = 61 7.0% 3 correct n = 136 15.5% 4 correct n = 184 20.8% 5 correct n = 204 23.3% 6 correct n = 184 21.0% 7 correct n = 83 9.5% Although only a few participants were completely accurate in their profile, the majority of the participants did get more than half of the responses correct. This does not provide support for the general hypothesis, which states that the level of inaccurate stereotyping will be high (3a) Instead, it seems as though most participants were able to provide a semi accurate offender profile Figure 4 2 Additive Responses for the Stereotypical Sex Offender Measures. 12 2

PAGE 123

. Figure 4 2 shows a bar chart of correct responses. The slightly skewed curve shows another view of the number of correct responses. Again, because the majority of participants correctly identified more than half of the measures, these findings do not provide support for the general hypothesis suggesting that the level of inaccurate stereotyping would be high. After the u nivariate analysis was completed, bivariate analyses were conducted between the Stereotypical Sex Offender variable and all of the control variables. Several variables age (r = .224) education level (r = .136) and socio economic status (r= .108) (all significant at the .001 alpha level) show a significant correlatio n with the Stereotypical Sex Offender variable. These correlations can be shown in Table B 22 (Appendix). Because these three significant correlations exist, the second hypothesis of the model is only partially supported (3b) That hypothesis states tha t the level of stereotyping will be similar across all social categories and groups. The last step in this analysis of the stereotypical sex offender was to conduct two independent sample t tests to determine if there was a significant difference between parents and non parents in the way they identified the stereot ypical sex offender. Table 4 12 shows the results of the two t tests. In t he first t test, there was not a significant difference parents (M= 4.46 SD= 1.624 ) and non parents (M= 4.65 SD= 1.420 ); t(875)= 1.750 p=. 080 for the total count of the stereotypical sex offender measures. Since there was not a significant difference between parents and non parents for th is t test the results do not provide support for the parent hypothesis in which parents overall predict the more accurate offender profile of the stereotypical sex offender, compared to the non parent participant group (3c) Second, a t test was run to compare 123

PAGE 124

gender against the total count of the stereotypical sex offender measures. In that t test, a significant difference was found between male participants (M= 4.27, SD=1. 539 ) and female participants (M= 4.86 S D= 1.475 ); t(875)= 5.816 p=. 000*** This t test suggests that women are more accurately able to identify the stereotypical sex offender in comparison to men. This goes against the hypothesis that suggested women would be more inaccurate than men (3d) These results are indicative that gender ma y be a predictive factor for many of the regression models in this project. Table 4 12 Independent Sample T Test for Parental Status or Gender and the Stereotypical Sex Offender Measures. Mean SD N df T Sig. Total Stereotypical Count Parents 4.46 1.624 479 875 1.750 080 Non Parents 4.65 1.420 398 875 1.750 Total Stereotypical Count Male 4.27 1.539 4 65 875 5.816 000*** Female 4.86 1.475 412 875 5.816 *** p<.001 **p < .01 *p < .05 Hypothesis Testing for Fear of Sexual Victimization Models Model 3: OLS Regression Model Predicting Fear of Victimization For Model 3, several hypotheses have been derived concerning the participants' fear of victimization. First the fear of personal victimization hypotheses will be addressed It is hypothesized that the level of fear of personal victimization will be moderate (4a) For this model it was hypothesized that the level of fear of personal victimization will be similar across social categories and groups (4b) Furthermore, non parents will report more fear of personal victimization than will parents (4c) and women will report more fear of personal victimization than will men (4d) Addressing the fear of victimi zation of children hypotheses, the level of fear of victimizati on of children will be high (5a) I t is also predicted that the level of fear will be similar across social categories and groups (5b) Additionally, parents will report more fear of victimization of 124

PAGE 125

children than will non parents (5c) and women will re port more fear of victimization of children than will men (5d) T wo dependent variables are used in this model 1) "Are you worried that children in your community may be at risk of becoming the victim of a sexual offense?" and 2) "Are you worried that y ou personally may be at risk of becoming the victim of a sexual offense?" Participants were asked to answer these questions based on a four point Likert scale with the responses of 1) "Definitely Not", 2) "Probably Not", 3) "Probably Yes" and 4) "Definite ly Yes". The results reveals that participants are largely split in their worry about children becoming victims of sexual offenses likely due to the parental status of the participants. This does not provide support for the fear of victimization of chi ldren, which hypothesizes that high levels of fear will be present (5a) In contrast, p articipants overall are not very worried (n = 610, 69.6% collectively) that they will become victims of a sexual offense. This finding somewhat supports the hypothesis that fear of personal victimization will be moderate (4a) Ta ble 4 13 Univariate for Fear of Victimization Responses. Measure Code Frequency Percent Are you worried that children in your community may be at risk of becoming the victim of a sexual offense? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes n = 188 n = 225 n = 288 n = 176 21.4% 25.7% 32.8% 20.1% Are you worried that you personally may be at risk of becoming the victim of a sexual offense? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes n = 256 n = 354 n = 181 n = 86 29.2% 40.4% 20.6% 9.8% N ext a bivariate correlation was conducted to examine the relationship between the control variables and the two types of fear of victim ization. First for the fear of personal victimization variable, only two control variables show significant correlations. 125

PAGE 126

Gender ( r = .068, significant at the .05 alpha level) and age ( r = .128, significant at the .001 alpha level) were the only two to produce a significant correlation with fear of personal victimization. For fear of personal victimization, these results are able to provide some support for the more specific hypotheses predicted for this model. First because two of the variables showed a signifi cant correlation, the second hypothesis regarding similarities of fear across all social categories and groups is only partially supported (4b) The parent hypothesis stating that non parents will have more personal fear of victimization compared to par ents is not supported (4c) Finally, the gender hypothesis which states that women will have more personal fear than men is supported (5d) Second for the fear of victimization for children variable, three control variables show a significant correlation. Parental status (r = .210, significant at the .001 alpha level) number of school age children ( r = .123, significant at the .001 alpha level) and age ( r = .089, significant at the .01 alpha level) provide significant correlations with fear of victimization of children. The various hypotheses regarding fear of victimization of children are supported at different levels. First, because three of the variables showed a significant correlation, the hypothesis regarding similarities of fear acro ss all social categories and groups is only partially supported (5b) Second, the parent hypothesis stating that parents will have more fear of victimization of children compared to non parents is supported (5c) Finally, the gender hypothesis whic h states that women will have more fear of victimization of children than men is also supported (5d) These correlations can be found in Table B 24. 126

PAGE 127

The next step in the analysis of this model was to conduct an independent sample t test to see if there was any significant difference in the fear of victimization measures between the parent and non parent group. The results of that t test (Table 4 17 ) show that there is a strong, significant difference between parents (M=2.71, SD=.959) and non parents (M= 2.28, SD=1.083); t(875)= 6.272, p=.000***, concerning the fear of children becoming victims of a sexual offense. This infers that the parents have a statistically higher fear of children being victimized compared to the non parent group something that p rovides further support to the parent hypothesis which theorizes exactly that relationship (5c) The mean difference between the parent and non parent responses also is reflective of parents having the larger fear that children will be victimized. The r esults of the second t test show that there is also a significant difference between male participants (M=2.43, SD=1.013) and female participants (M=2.62, SD=1.059); t(875)= 5.880, p=.006**, concerning fear of children becoming victims of a sexual offense. These results suggest that female participants are more afraid then the male participants of children becoming victims of a sexual offense (5d) This makes sense given the significant difference in parents and non parents and their fear of children bein g victimized. Many of the females will also identify as being mothers, keeping a consistent relationship between the two t tests. The results of the third t test show that there is no significant difference between parents (M=2.16, SD=.918) and non parent s (M=2.05, SD=.959); t(875)= 1.732, p=.552, concerning fear of personal victimization. This infers that both groups have similar fear levels of being victimized, and from the results of the frequency analysis neither group is overly concerned that victimi zation is a realistic possibility. The results of the 127

PAGE 128

independent sample t test further do not support the parent hypothesis which states that the non parent participants would be more afraid of being personally victimized than the parent participants (4 c) The small mean difference between parents and non parents is also reflective of the two participant groups having similar fear levels for personal victimization. Instead, the results show that neither group has very high levels of fear regarding pers onal victimization. The results of the fourth t test show that there is a significant difference between male participants (M=1.94, SD=.921) and female participants (M=2.31, SD=.919); t(875)= 2.743, p=.000* **, concerning fear of personal victimization. T his suggests that females are more afraid of personally becoming victims of a sex offense in comparison to makes (4d) this finding follows prior research findings which also suggests a gender difference regarding fear of crime. The difference in gender for both t tests in this analysis fits with the prior t tests on accurate knowledge and the stereotypical sex offender measures. Based on these findings, gender is showing signs of being a significant predictor for the remainder of the analysis to be run. Table 4 14 Independent Sample T Test for Parental Status or Gender and Fear of Victimization Victimization of Children Mean SD N df T Sig. Parents 2.71 .959 479 875 6.272 .000*** Non Parents 2.28 1.083 398 875 6.272 Victimization of Children N df T Sig. Mean SD Males 2. 43 1.013 4 64 875 2.743 .006 ** Females 2.63 1.059 412 875 2.743 Personal Victimization Parents 2.16 .918 479 875 1.732 .552 Non Parents 2.05 .959 398 875 1.732 Personal Victimization N df T Sig. Mean SD Males 1.94 .9 21 4 64 875 5.880 .000*** Females 2.31 .919 412 875 5.880 *** p<.001 **p < .01 *p < .05 128

PAGE 129

In the last set of analyses for this model, two OLS regression models were run to look at the linear relationship between parental status and the fear of vi c timization measures. Table 4 15 shows the OLS regression model predicting the participant's fear of victimization The additive registry knowledge scale served as the independent variable, along with the control variables of the stere otypical offender, parental status, number of school age children, gender, age, race, ethnicity, education, marital status, income level, socio economic status (SES) geographic region, and population size the frequency measured for these control variable s can be found in Tables 4 1 through 4 4. The first model (Fear of Victimization of Children) was strongly significant at the .001 alpha level, and the predictor vari ables are able to account for 9 9 % of the explained variance, as shown by the R Square. There are four variables that are significant within the model the stereotypical offender (significant at the .01 alpha level), parental status and age (both significant at the .001 alpha level), and race (significant at the .01 alpha level). There is n o significant relationship between registry knowledge and fear. Additionally, the relationship between the two variables is negative indicating that as the level of knowledge decreases, the level of fear of victimization of children increases Furtherm ore since the parental status variable is significant, this regression analysis provides further support for the parent hypothesis which predicted that parents would be more fearful of children being victimized comparative to the non parent group. Since gender is not significant for the fear of victimization of children model, this particular finding does not support the hypothesis that stated women would be more fearful than men regarding this issue. 129

PAGE 130

For the secon d regression analysis, Tab le 4 15 shows the OLS regression model predicting the participant's fear of personally being victimized. The additive registry knowledge scale served as the independent variable, along with the same control variables used in the previous regression model. The o verall model was strongly significant at the .001 alpha level, and the predictor varia bles are able to account for 9.0 % of the explained variance, as shown by the R Square. There are four variables that are significant within the model parental status a nd gender ( both significant at the .05 alpha level), age (significant at the .01 alpha level) and race (significant at the .0 1 alpha level). The negative relationship between registry knowledge and fear of personal victimization indicates that when partic ipants exhibit more inaccurate levels of registry knowledge, they will also exhibit more fear of personally becoming a victim of a sexual offense. Interestingly, both age and race show a negative relationship with personal victimization. Based on the var iables, this suggests that older individuals and white participants are less fearful of becoming victims of sexual offenses. Due to the positive and significant relationship found between the parental status variable and the fear of personal victimization the results of this regression fur ther do not support the parent which states that the non parent participants would be more afraid of being personally victimized than the parent participants. If the hypothesis had been supported, then a negative relat ionship would exi st between those two variables. However, because gender is a significant variable for the fear of personal victimization model, this suggests that the gender hypothesis is supported which stated that women would be more personally afrai d of being victimized compared to men. These results are consistent with previous literature (Ferraro, 1995). 130

PAGE 131

Table 4 15 OLS Regression Predicting Fear of Children Being Victimized and Fear of Personal Victimization Fear of Children Being Victimized Fear of Personal Victimization Variable Coefficient Std. Error Beta Coefficient Std. Error Beta Registry Knowledge .017 .016 .034 .024 .014 .055 Stereotypical Offender .063 ** .025 .094 .015 .022 .025 Parental Status .598 *** .087 .287 .192* .079 .102 Num. School Age Children .011 .046 .009 .009 .042 .008 Gender .016 .010 .052 .018* .009 .064 Age .085 *** .017 .1 9 4 .044 ** .015 .113 Race .233 ** .078 .110 .389 *** .071 .204 Ethnicity .007 .007 .032 .009 .006 .045 Education Level .030 .074 .014 .049 .067 .026 Marital Status .075 .073 .035 .001 .066 .000 Income Level .074 .049 .056 .030 .045 .024 SES Status .017 .067 .007 .101 .061 .062 Geographic Region .030 .071 .015 .006 .064 .003 Population Size .007 .027 .008 .002 .025 .002 Constant 2.325 .192 2.448 .174 F Statistic 6.729 *** 6.106 *** R Square 099 .090 *** p<.001 **p < .01 *p < .05 Parental Status (dummy coded: Non Parent/Parent), Race (dummy coded: Non White/White), Ethnicity (dummy coded: Non Hispanic/Hispanic), Education Level (dummy coded: No Bachelor's Degree/Bachelor's Degree), Marital Status (dummy coded: Not Married/Not Married), Geographic Region (Not from the South/From the South) Hypothesi s Testing for Perceptions of Community Related Attitudes Models Factor Analysis for Moral Panic Measures Before delving into the regression models to examine the five elements of the predicted moral panic, a final preliminary analysis must be completed to make sure that the measures are reliable to use in the regressions. Table 4 16 shows that the twenty five measures used to examine a moral panic, load onto five different components the exact number that was desired based on Cohen's original conceptual ization. With only three exceptions, all of the measures load onto the correct component. 131

PAGE 132

Table 4 16 Factor Analysis for Moral Panic Measures. Component 1 2 3 4 5 Are you worried about sex offenders living nearby your home? (Concern) .557 Are you worried that children in your community may be at risk of becoming the victim of a sexual offense? (Concern) .903 Are you worried that you personally may be at risk of becoming the victim of a sexual offense? (Concern) .615 Are you worried about children in your community being at risk of being approached by a sexual offender? (Concern) .873 Are you worried that as sex offenders continue to live in the community, then more sex offenses will occur? (Concern) .541 Are you angry that sex offenders are allowed to live in the community? (Hostility) .789 Do you feel any resentment over the fact that some of your neighbors may be sex offenders? (Hostility) .676 Do you feel any anger towards the criminal justice system for releasing sex offenders from jails and prisons? (Hostility) .746 Are you angry that sex offenders may be working at businesses where you may frequently shop or visit? (Hostility) .809 Are you angry that children in your community might come into contact with a sex offender? (Hostility) .571 Do you think that a majority of community members are in agreement about the risk that sex offenders pose? (Consensus) .447 Do you think that many community members feel that changes must be made in the supervision of sex offenders? (Consensus) .642 Do you that the community members in general feel threatened by sex offenders as a group? (Consensus) .794 Do you think that a majority of community members are in agreement that children are at risk of being sexually victimized? (Consensus) .767 Do you think that many community members feel that sex offenders are too dangerous to be living in the community? (Consensus) .773 Do you feel that the current state of the sex offender registry is too harsh?* (Disproportionality) .702 Do you feel that the sex offender registry laws should be stricter? (Disproportionality) .634 Do you think that keeping sex offenders on electronic monitoring/GPS tracking for more than 5 years without a break is too severe a punishment?* (Disproportionality) .722 Do you think that sex offenders should report to law enforcement more than the required two times per year? (Disproportionality) .511 Do you think that the media overreacts in their reporting of sex offenses when they occur in a community?* (Disproportionality) .786 132

PAGE 133

Table 4 16. Continued Component Do you think that law enforcement reacts quickly when a sexual offense takes place? (Volatility) .846 Do you think that legislators work fast enough to get necessary registry laws passed to further keep track of sex offenders? (Volatility) .691 Do you think that the media reports sex offense cases too quickly before all of the facts are gathered? (Volatility) .653 Do you think that the quick response of the media makes communities safer because people are made aware of the sex offense? (Volatility) .470 Do you think that police are too slow to catch sex offenders when sex offenses take place? (Volatility) .601 Indicates that the measures were reverse coded Cronbach's Alpha (Offender Measures) = .841 (Concern), .885 (Hostility), .796 (Consensus), .785 (Disproportionality) .645 (Volatility). Cronbach's Alpha (Predator Measures) = .846 (Concern), .890 (Hostility), .813 (Consensus), .788 (Disproportionality), .652 (Volatility). This is an improvement from the pilot study data. In the pilot study, the factor analysis still produced five components but the individual measures did not load as neatly. For the full study, several measures were reworded and extra measures were added so that all five moral panic elements have five measures each. After doing so, most of the measures load onto the component factors at a higher level then they did before. Furt hermore, a reliability analysis was completed on each of the different components. Finally because nearly identical measures were used for the offender and predator randomization, the factor analysis should be no different for the predator randomization then it was for the offender randomization. To confirm this thought, a second reliability analysis was completed for the predator measures the Cronbach's Alphas for the predator measures were nearly identical to those of the offender measures. This con firms that the measures are reliable for both portions of the randomization. 133

PAGE 134

Model 4: OLS Regression Model Predicting the Moral Panic Element of Concern For Model 4, four hypotheses have been derived concerning the participants' level of concern regarding sex offenders and the threat they pose to the community. First, concern about sexual offenses and offending will be high (6a) Second, will be similar across social categories and groups, and it will be similar for the group responding to "sex offender" a nd that responding to "sex predator" survey items (6b) Third parents will be more concerned than non parents (6c) Fourth women will be more concerned than men (6d) This set of questions used the implemented random assignment of "sexual offender" and "sexual predator" questions to the participants. In order to test this hypothesis, frequency analysis, bivariate analysis, independent sample t tests and OLS Regressions will be completed. Table 4 17 shows the u nivariate analysis for the moral panic element of concern regarding sex offenders living in our communities. This was one of the measures that implemented the random assignment of "sex ual offender" and "sex ual p redator" questions. The results show that for the most part, participants are largely split in the middle of all of the questions. The decision was made not to use a 5 point Likert scale with an "Unsure" option, in order to somewhat force participants in to the direction of a Yes or No answer. The results show that participants are largely clustered around the "Probably Not" or "Probably Yes" options, which still indicates some uncertainty in their responses but still leads them in one direction or the ot her. Since there was no definitive response across participants, the first hypothesis stating that concern will be high cannot be supported at this time (6a) 134

PAGE 135

Table 4 17 Univariate Analysis for Moral Panic Element of Concern (Offender and Predator Randomization). Measure (Offender n = 452) Code Frequency Percent Are you worried about sex offenders living nearby your home? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) n = 55 n = 135 n = 148 n = 114 n = 425 6.3% 15.4% 16.8% 13.0% 48.5% (51.5%) (100.0%) Are you worried that children in your community may be at risk of becoming the victim of a sexual offense? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) n = 89 n = 110 n = 165 n = 88 n = 425 10.1% 12.5% 18.9% 10.0% 48.5% (51.5%) (100.0%) Are you worried that you personally may be at risk of becoming the victim of a sexual offense? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) n = 131 n = 188 n = 86 n = 47 n = 425 14.9% 21.4% 9.8% 5.4% 48.5% (51.5%) (100.0%) Are you worried about children in your community being at risk of being approached by a sexual offender? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) n = 83 n = 100 n = 167 n = 102 n = 425 9.5% 11.4% 19.0% 11.6% 48.5% (51.5%) (100.0%) Are you worried that as sex offenders continue to live in the community, then more sex offenses will occur? Measure (Predator n = 425) 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) Code n = 37 n = 111 n = 195 n = 109 n = 425 Frequency 4.2% 12.7% 22.2% 12.4% 48.5% (51.5%) (100.0%) Percent Are you worried about sex predators living nearby your home? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 52 n = 140 n = 142 n = 91 n = 452 5.9% 16.0% 16.2% 10.4% 51.5% (48.5%) (100.0%) Are you worried that children in your community may be at risk of becoming the victim of a sexual offense? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 97 n = 116 n = 119 n = 93 n = 452 11.1% 13.2% 13.6% 10.6% 51.5% (48.5%) (100.0%) 135

PAGE 136

Table 4 17 Continued. Measure (Offender n = 452) Code Frequency Percent Are you worried that you personally may be at risk of becoming the victim of a sexual offense? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 129 n = 163 n = 92 n = 41 n = 452 14.7% 18.6% 10.5% 4.7% 51.5% (48.5%) (100.0%) Are you worried about children in your community being at risk of being approached by a sexual predator? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 96 n = 114 n = 117 n = 98 n = 452 10.9% 13.1% 13.3% 11.2% 51.5% (48.5%) (100.0%) Are you worried that as sex predators continue to live in the community, then more sex offenses will occur? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 46 n = 114 n = 176 n = 89 n = 452 5.2% 13.2% 20.0% 10.1% 51.5% (48.5%) (100.0%) Two scales were created from these five measures one was the scale variable for the participant's concern regarding sex offenders, and the other regards sex predators. The two scales were created by adding the scores of the five measures acr oss participants, and then by dividing that score by five. Once the scales were completed then bivariate correlations were conducted to see if there were any significant correlations between the concern scales and the control variables. Since there was a random assignment of "sexual offender" and "sexual predator" implemented for the concern measures, the bivariate correlations were conducted with that random assignment in mind. In the first set of biv ariate correlations (Table B 26, Appendix), three control variables show a significant correlation with the Concern scale (Offender sub sample ). Parental status ( r = .195, significant at the .001 alpha level), number of school age children ( r = .144, significant at the .01 alpha level) and gender ( r = .103, 136

PAGE 137

significant at the .05 alpha level) are provide significant correlations with Concern. The various hypotheses regarding concern are supported at different levels. First, because three of the varia bles showed a significant correlation, the hypothesis regarding similarities of concern across all social categories and groups is only partially supported (6b) Second, the parent hypothesis stating that parents will be more concerned compared to non p arents is supported (6c) Finally, the gender hypothesis which states that women will be more concerned than men is also supported (6d) In the second bivariate correlation (Table B 27), the same three variables provided a correlation with the Con cern scale (Predator sub sample ). Parental status ( r = .192, significant at the .001 alpha level), number of school age children ( r = .115, significant at the .01 alpha level) and gender ( r = .187, significant at the .001 alpha level) are provide signific ant correlations with Concern (Predator sub sample ). Since the hypotheses are the same for both the offender and the predator randomization, they are all supported in the same way they were for the offender model. Next, two independent sample t tests were conducted to see if there is any significant difference between the parent and non parent groups these scaled variables. Table 4 18 shows the two t tests that there is a moderately significant difference between the parents (M=2.42, SD=.79) and non paren ts (M=2.71, SD=.69); t(450 )= 4.088, p=.006** for the offender randomization. There was also a moderately significant difference between parents (M=2.65, SD=.81) and non parents (M=2.35, SD=.70); t(423 )= 3.96, p=.008** for the predator randomization. This shows that parents are more concerned about sexual offenders/predators and the subsequent risk they pose in 137

PAGE 138

comparison to their non parent counterparts, further supporting the parent hypothesis (6c) For the gender t tests, the results show that there is a moderately significant difference between the male (M=2.47, SD=.73 ) and female participants (M=2.70 SD=. 75 ); t(450 )= 3.209, p=.001 ** for the offender randomization. There was also a strongly significant di fference between parents (M=2.38, SD=.74 ) and n on parents (M=2.67, SD=.77); t(423 )= 3. 851, p=.000 ** for the predator randomization Both of these t tests suggest that women are more concerned about the threat that sex offenders pose compared to men (6d) This finding is consistent with the fear of cr ime literature that suggests that concern is a common reaction to crime, and even victimization (Ferraro and LaGrange, 1987). Table 4 18 Independent Sample T Test for Parental Status and the Element of Concern Offender Randomization Mean SD N df T Sig. Parents 2.42 .79 250 450 4.088 .006** Non Parents 2.71 .69 202 450 4.088 Offender Randomization Male Female 2.47 2.70 .73 .75 237 214 450 450 3.209 3.209 .001*** Predator Randomization Parents 2.65 .81 228 423 3.96 0 .008** Non Parents 2.35 .70 197 423 3.96 0 Predator Randomization Male Female 2.38 2.67 .74 .77 227 198 423 423 3.851 3.851 .000*** *** p<.001 **p < .01 *p < .05 The next step in analyzing this model is to run two OLS regression to predict the moral panic element of concern. One model will be run for the offender randomization and the other for the predator randomization. The additive registry knowledge variable will still serve as the independent variable and all of the previously used controls wil l remai n control variables. Table 4 19 shows both of those OLS Regressions. Overall, 138

PAGE 139

both models are strongly significant at the .001 alpha level and the predictor variables are able to account for 11.9 % (offender) and 14.9 % (predator) of the respective explained variance, as shown by the R Squares. For the offender model there are four variables that are significant within the model parental status (significant at the .001 alpha level), gender (significant at the .05 alpha level), age (significant at the .01 alpha level) and race (significant at the .001 alpha level). For the predator model, there are five variables that are significant within the model registry knowledge (significant at the .01 alpha level), parental status, gender (both significan t at the .001 alpha level), age and race (both significant at the .01 alpha level). For both models, there is a negative relationship present for age and race, which suggests that younger individuals and minority participants are more concerned about sex offenders and the risks that they pose. Due to the positive and significant relationship found between the parental status variable and concern, the results of this regression furthe r partially support the parent hypothesis, which states that parent part icipants would be more concerned about sex offenders compared to non parents The registry knowledge variable did not significant predict concern for the offender randomization, but did so for the predator one. This is an interesting finding in itself si nce it provides support for the notion that there is something systematically different about the terms "sex ual offender" and "sex ual predator," making the latter more salient in the minds of the participants this does not p rovides support for the first hypothesis, which stated there would be no difference regarding measures involving "sexual offenders" and "sexual predators" Furthermore, the negative relationship that exists between registry knowledge and concern within the predator model, suggests tha t as 139

PAGE 140

the level of knowledge de creases among participants, so does their concern. Gender is significant in both the offender and predator models, but it is stronger for the predator model. This supports the gender hypotheses, which predicted that women wo uld be more concerned than men. This mixture of results provides initial support for the perceived presence of a moral panic regarding sex offenses, sex offenders and the registry, which supervises them. The next four models predict the other elements of Cohen's moral panic. T able 4 19 OLS Regression Predicting Element of Concern (Random Assignment). Variable Offender Randomization Predator Randomization Coefficient Std. Error Beta Coefficient Std. Error Beta Registry Knowledge .016 .017 .044 .038 .018 .099 Stereotypical Offender .034 .025 .066 .048 .026 .097 Parental Status .330 *** .092 .219 .380 *** .091 .240 Num. School Age Children .036 .046 .042 .024 .048 .025 Gender .014* .007 .090 288 *** .074 188 Age .043** .017 .139 .056** .018 .173 Race .312*** .083 .200 .231** .083 .150 Ethnicity .117 .112 .050 .002 .005 .016 Education Level .031 .075 .020 .028 .080 .018 Marital Status .052 .076 .035 .035 .077 .023 Income Level .053 .048 .057 .030 .057 .028 SES Status .026 .067 .021 .101 .075 .073 Geographic Region .102 .072 .066 .106 .075 .067 Population Size .029 .028 .049 .023 .029 .039 Constant 2.657 .330 F Statistic 4.077 *** 4.970 *** R Square .119 .149 *** p< .001 **p < .01 *p < .05 Parental Status (dummy coded: Non Parent/Parent), Race (dummy coded: Non White/White), Ethnicity (dummy coded: Non Hispanic/Hispanic), Education Level (dummy coded: No Bachelor's Degree/Bachelor's Degree), Marital Status (dummy code d: Not Married/Not Married), Geographic Region (Not from the South/From the South) 140

PAGE 141

Model 5: OLS Regression Predicting the Moral Panic Element of Hostility For Model 5, several hypotheses have been derived concerning the participants' level of hostility regarding sex offenders and the threat they pose to the community. This set of questions used the implemented random assignment of "sexual offender" and "sexual predator" questi ons to the participants. First, the hostility/anger about sex offenses and offenders will be high (7a) Second, the level of hostility/anger will be similar across social categories and groups, and it will be similar for the group responding to "sexual o ffender" and that responding to "sexual predator" survey items (7b) Third parents will be more hostile/angry than non parents (7c) Finally, women will be more hostile/angry than men (7d) In order to test this hypothesis, frequency analysis, bivariat e analysis, independent sample t tests and OLS Regressions will be completed. Table 4 20 7 shows the u nivariate analysis for the moral panic element of hostility against sex offenders living in our communities. Five measures were used to measure hostility and were included in the implemented random assignment of "sex ual offender" and "sex ual predator." The hostility variables were measured just like the concern variables. The results show that for the most part, participants clustered toward the middle of all of the questions, but there is a slight indication that participants are responding with more anger meaning the answers are more frequently in the "Probably Yes" and "Definitely Yes" categories rather than the negative options. 7 The table reflects the measures for both the offender and the predator random condition measures. The table breaks up percentages for the whole sample so that within the offender or predator c onditions, the reported figures do not add up to 100%. 141

PAGE 142

Table 4 20 Univari ate Analysis for Moral Panic Element of Hostility (Offender and Predator Randomization). Measure (Offender n = 452) Code Frequency Percent Are you angry that sex offenders are allowed to live in the community? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) n = 53 n = 140 n = 163 n = 96 n = 425 6.0% 16.0% 18.5% 11.0% 48.5% (51.5%) (100.0%) Do you feel any resentment over the fact that some of your neighbors may be sex offenders? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) n = 49 n = 131 n = 171 n = 101 n = 425 5.6% 14.9% 19.5% 11.5% 48.5% (51.5%) (100.0%) Do you feel any anger towards the criminal justice system for releasing sex offenders from jails and prisons? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) n = 61 n = 128 n = 168 n = 95 n = 425 7.0% 14.5% 19.2% 10.8% 48.5% (51.5%) (100.0%) Are you angry that sex offenders may be working at businesses where you may frequently shop or visit? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) n = 64 n = 152 n = 149 n = 87 n = 425 7.3% 17.2% 17.0% 10.0% 48.5% (51.5%) (100.0%) Are you angry that children in your community might come into contact with a sex offender? Measure (Predator n = 425) 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) Code n = 52 n = 89 n = 161 n = 150 n = 425 Frequency 5.9% 10.1% 18.4% 17.1% 48.5% (51.5%) (100.0%) Percent Are you angry that sex predators are allowed to live in the community? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 52 n = 134 n = 159 n = 80 n = 452 5.9% 15.3% 18.2% 9.1% 51.5% (48.5%) (100.0%) Do you feel any resentment over the fact that some of your neighbors may be sex predators? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 48 n = 124 n = 174 n = 79 n = 452 5.5% 14.1% 19.8% 9.1% 51.5% (48.5%) (100.0%) 142

PAGE 143

Table 4 20 Continued Measure (Offender n = 452) Code Frequency Percent Do you feel any anger towards the criminal justice system for releasing sex predators from jails and prisons? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 57 n = 131 n = 144 n = 93 n = 452 6.6% 14.9% 16.4% 10.6% 51.5% (48.5%) (100.0%) Are you angry that sex predators may be working at businesses where you may frequently shop or visit? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 62 n = 139 n = 141 n = 83 n = 452 7.1% 15.8% 16.1% 9.5% 51.5% (48.5%) (100.0%) Are you angry that children in your community might come into contact with a sex predator? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 57 n = 101 n = 140 n = 127 n = 452 6.5% 11.5% 16.0% 14.5% 51.5% (48.5%) (100.0%) These are the first indications that participants do have some hostile feelings or feelings of anger directed toward sex offen ders living back in communities, and provides initial support for the first hypothesis, which suggests that hostility and anger toward sex offenders will be high (7a). The next step in this analysis was to create two scales from the five offender and five predator measures. The two scales were created by adding the scores of the five measures across participants and then divide that score by five. Once the scales were completed then, bivariate correlations were run to determine if hostility was correlate d with any of the control variables. Since there was a random assignment of "sexual offender" and "sexual predator" implemented for the hostility measures, the bivariate correlations were conducted with that random assignment in mind. 143

PAGE 144

In the first set o f bivariate correlations (Table B 28, Appendix), two control variables show a significant correlation with the Hostility scale (Offender sub sample ). Parental status ( r = .196, significant at the .001 alpha level), and number of school age children ( r = 113, significant at the .01 alpha level) both provide significant correlations with Hostility The various hypotheses regarding hostility are supported at different levels. First, because only two of the variables showed a significant correlation, the hy pothesis regarding similarities of hostility across all social categories and groups is only partially supported (7b) Second, the parent hypothesis stating that parents will be more hostile/angry compared to non parents is supported (7c) Finally, t he gender hypothesis which states that women will be more hostile/angry than men is not supported because there is so significant correlation (7d) For the second set of bivariate correlations (Table B 29, Appendix), three control variables show a significant correlation with the Hostility scale (Predator sub sample ). Parental status ( r = .194, significant at the .001 alpha level), and number of school age children ( r = .115, significant at the .01 alpha level), and gender ( r = .201, significant at the .001 alpha level) all provide significant correlations with Hostility. Because the same hypotheses are predicted for both the offender and predator models, the same support is being exhibited in the second bivariate correlation, with one exception. The gender hypothesis that women will be more concerned than men was not supported in the offender model, but it is supported for the predator model (7d) Next, two independent sample t tests were conducted to see if there is any significant difference between the parent and non parent groups for the se scaled variables. Table 4 21 shows the results of the two t tests. The first shows that there is 144

PAGE 145

no significant difference in hostility between the parents (M=2.85, SD=.80) and non parents (M=2.54, SD=.74); t(450 )= 4.176, p=.069 for the participants assigned to the offender language randomization. There was a moderate significant difference in hostility between parents (M=2.80, SD=.72) and non parents (M=2.50, SD=.83); t(423 )= 3.976, p=.003** for t he participants assigned to the predator language randomization. This shows that parents are angrier about sexual offenders or predators living in the community, further supporting the parent hypothesis (7c) However, there is a stronger level of signifi cance for the second t test, indicating that the participants are directing more of their hostility toward sex predators compared to sex offenders, which contradicts the hypothesis, which states that offender and predator measures will be answered similarl y (7b) Finally, t he second shows that there is a moderately significant difference in hostility between the male (M=2.60, SD=.72 ) and female participants ( M=2.82 SD=.83); t(450)= 2.903, p=.004** for the participants assigned to the offender language rand omization. There was a strongly significant difference between male (M=2.51, SD=.79 ) and female participants (M=2.83, SD=.76); t(423 )= 4.156, p=.000* ** for the participants assigned to the predator language randomization. These two t tests both suggest t hat females are more hostile/angry in concerning sex offenders. These results support the gender hypothesis, which suggests this relationship (7d) Furthermore, the predator t test is more strongly significant than the offender t test suggesting that par ticipants are able to tell the difference between the two terms. Although this does not support the hy pothesis regarding the offender or predator measures, it does provide good evidence that participants can break down the nuances between the terms. 145

PAGE 146

Table 4 21 Independent Sample T Test for Parental Status and the Element of Hostility Offender Randomization Mean SD N df T Sig. Parents 2.85 .80 250 450 4.176 .069 Non Parents 2.54 .74 202 450 4.176 Offender Randomization Males Females 2.60 2.82 .72 .83 237 214 450 450 2.903 2.903 .004** Predator Randomization Parents 2.80 .72 227 423 3.976 .003** Non Parents 2.50 .83 198 423 3.016 Predator Randomization Males Females 2.51 2.83 .79 .76 227 198 423 423 4.156 4.156 .000*** *** p<.001 **p < .01 *p < .05 The next step in analyzing this model is to run two OLS regression to predict the moral panic element of hostility One model will be run for the offender randomization and the other for the predator randomization. The additive registry knowledge variable will still serve as the independent variable and all of the previously used controls will remai n control variables. Table 4 22 shows both of those OLS Regressions. Overall both models are significant the offender re gression is significant at the .01 alpha level and the predator regression is significant at the .001 alpha level. The predictor varia bles are able to account for 6.7 % (offender) and 13 .5 % (predator) of the respective explained variance, as shown by the R Squares. For the offender model only the parental status and race variables (significant at the .01 alpha level) are significant predictors for participant hostility. This finding is consistent with the results of the t tests, which showed a significant difference between the two parent participant groups. For the second regression model, registry knowledge (significant at the .05 alpha level) parental status (significant at the .01 level), gender and race (significant at the .001 level) are all signifi cant predictor variables for the participant's hostility. Many of these predictor variables are not significant for the offender model, suggesting that sex predators are a more concerning grou p comparative to sex offenders. Due to the 146

PAGE 147

positive and signif icant relationship found between the parental status variable and the element of hostility, the results of this regression further partially support the parent hypothesis for this model which states that parent participants would have stronger feelings of hostility or anger against sex offenders in comparison to non parents Table 4 22 OLS Regression Predicting Element of Hostility (Random Assignment). Variable Offender Randomization Predator Randomization Coefficient Std. Error Beta Coefficient Std. Error Beta Registry Knowledge 01 3 .018 .034 .043* .01 9 .108 Stereotypical Offender .035 .027 .069 .050 .027 .097 Parental Status 305 ** .099 .190 .269** .095 .169 Num. School Age Children .021 .050 .023 .034 .050 .036 Gender .009 .008 .055 .290 *** .077 .197 Age .004 .018 .014 .008 .018 .023 Race .234 ** .089 .137 .302* ** .086 .176 Ethnicity .087 .121 .035 .005 .005 .040 Education Level .007 .080 .004 .079 .083 .049 Marital Status .039 .082 .025 .005 .080 .003 Income Level .035 .051 .036 .075 .059 .068 SES Status .004 .072 .003 .083 .078 .058 Geographic Region .095 .078 .058 .063 .078 .038 Population Size .018 .030 .029 .040 .030 .064 Constant 2.5 62 205 1.584 .330 F Statistic 2.17 9** 4.414 *** R Square .067 13 5 *** p<.001 **p < .01 *p < .05 Parental Status (dummy coded: Non Parent/Parent), Race (dummy coded: Non White/White), Ethnicity (dummy coded: Non Hispanic/Hispanic), Education Level (dummy coded: No Bachelor's Degree/Bachelor's Degree), Marital Status (dummy coded: Not Married/Not Married), Geographic Region (Not from the South/From the South) The registry knowledge variable did not significant predict hostility for the offender randomization, but did so for the predator one. There is a negative relationship between registry knowledge and hostility, which suggests that as levels of knowledge decrease, the participants' levels of hostility increase. Overall, the predator regression is stronger and able to explain more of the variance within the model. This is an 147

PAGE 148

interesting finding in itself since it provides support for the notion that there is something systematically different about the terms "sex ual offender" and "sex ual predator," making the latter more salient in the minds of the participants this does not however provide support for the initial hypothesis, which predicted that participants would not respond differently to the two terms. Gender is only a significant predictor variable for the predator regression and not for the offender regression. This provides partial support for the gender hypothesis in this model, which suggested that women would be more hostile/angry toward sex offenders compared to men. This further provides evidence that the predator model is the stronger of the two models. Model 6: OLS Regression Model Predicting the Moral Panic Element of Consensus For Model 6, several hypotheses have been derived concerning the participants' ex pressed level of consensus against sex offenders and the threat they pose to the community. This set of questions used the implemented random assignment of "sexual offender" and "sexual predator" questions to the participants. First, the consensus about sex offenses and offending will be high (8a) Then it is predicted that consensus will extend across social categories and groups, and it will be similar for the group responding to "sex offender" and that responding to "sex predator" survey items (8b) It is also predicted that parents will be more unified than non parents (8c) and that women will be more unified than men (8d) In order to test these hypothese s, frequency analysis, bivariate analysis, independent sample t tests and OLS Regressions will be completed. 148

PAGE 149

Table 4 23 Univariate Analysis for Moral Panic Element of Consensus (Offender and Predator Randomization). Measure (Offender n = 452) Code Frequency Percent Do you think that a majority of community members are in agreement about the risk that sex offenders pose? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) n = 35 n = 105 n = 188 n = 124 n = 425 4.0% 12.0% 21.4% 14.1% 48.5% (51.5%) (100.0%) Do you think that many community members feel that changes must be made in the supervision of sex offenders? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) n = 14 n = 64 n = 257 n = 117 n = 425 1.6% 7.3% 29.3% 13.3% 48.5% (51.5%) (100.0%) Do you that the community members in general feel threatened by sex offenders as a group? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) n = 20 n = 68 n = 192 n = 172 n = 425 2.3% 7.9% 21.8% 19.5% 48.5% (51.5%) (100.0%) Do you think that a majority of community members are in agreement that children are at risk of being sexually victimized? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) n = 12 n = 73 n = 249 n = 118 n = 425 1.4% 8.3% 28.4% 13.4% 48.5% (51.5%) (100.0%) Do you think that many community members feel that sex offenders are too dangerous to be living in the community? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) n = 13 n = 63 n = 241 n = 135 n = 425 1.5% 7.2% 27.5% 15.4% 48.5% (51.5%) (100.0%) Measure (Predator n = 425) Code Frequency Percent Do you think that a majority of community members are in agreement about the risk that sex predators pose? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 31 n = 93 n = 195 n = 106 n = 452 3.5% 10.6% 22.3% 12.1% 51.5% (48.5%) (100.0%) Do you think that many community members feel that changes must be made in the supervision of sex predators? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 13 n = 77 n = 219 n = 116 n = 452 1.5% 8.8% 25.0% 13.2% 51.5% (48.5%) (100.0%) 149

PAGE 150

Table 4 23 Continued. Measure (Offender n = 452) Code Frequency Percent Do you that the community members in general feel threatened by sex predators as a group? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 9 n = 72 n = 197 n = 147 n = 452 1.0% 8.2% 22.6% 16.7% 51.5% (48.5%) (100.0%) Do you think that a majority of community members are in agreement that children are at risk of being sexually victimized? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 11 n = 90 n = 204 n = 120 n = 452 1.3% 10.3% 23.3% 13.6% 51.5% (48.5%) (100.0%) Do you think that many community members feel that sex predators are too dangerous to be living in the community? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 10 n = 74 n = 207 n = 134 n = 452 1.1% 8.4% 23.6% 15.4% 51.5% (48.5%) (100.0%) Table 4 23 shows the univariate analysis for the moral panic element of consensus against sex offenders living in our communities. Five measures were used to measure consensus and were included in the implemented random assignment of "sex offender" and "sex predator." The consensus variables w ere measured just like the concern and hostility variables. The results show that for the most part, participants are responding in a more unified manner meaning the answers are more frequently in the "Probably Yes" and "Definitely Yes" categories rathe r than the negative options, indicating higher levels of consensus. This is the first indication that participants exhibit a collective response against sex offenders and sex predators, and the threat that they pose to community members. For these measur es, since the majority of participants answer in the positive direction (Yes over No), then the first hypothesis can be supported stating that consensus will be high (8a). 150

PAGE 151

The next step in this analysis was to create two scales from the five offender and f ive predator measures. The two scales were created by adding the scores of the five measures across participants and then divide that score by five. Once the scales were completed then, two sets of bivariate correlations can be run. Since there was a ra ndom assignment of "sexual offender" and "sexual predator" implemented for the consensus measures, the bivariate correlations were conducted with that random assignment in mind. In the first set of b ivariate correlations (Table B 30 Appendix), only age ( r = .131, significant at the .01 alpha level) proves to have a significant correlation with Consensus (Offender sub sample ). In the second set of bivariate correlations (Table B 31, Appendix) three variables exhibit a significant correlation with Consens us (Predator sub sample ). Number of school age children ( r = .108, significant at the .05 alpha level), gender ( r = .211, significant at the .001 alpha level) and age ( r = .104, significant at the .05 alpha level) are all significant predictors of Consens us (Predator sub sample ). The various hypotheses regarding consensus are supported at different levels. First, because one variable is significant for the offender model and three of the variables (other than parent status and gender) showed a significa nt correlation for the predator model the hypothesis regarding similarities of consensus across all social categories and groups is not well supported (8b) Second, the parent hypothesis stating that parents will be more united than non parents is no t supported for either model (8c) Finally, the gender hypothesis which states that women will be more united than men is supported but only for the predator model (8d) 151

PAGE 152

Next, two independent sample t tests were conducted to see if there is any sig nificant difference in consensus between the parent and non parent groups for these scaled variables. Table 4 24 shows the results of the two t tests. The first shows that there is a weak significant difference in consensus between the parents (M=3.08, SD=.59) and non parents (M=3.04, SD=.53); t(450 )= .770, p=.020* for the participants assigned to the offender language randomization There was no significant difference between in consensus parents (M=3.06, SD=.59) and non p arents (M=3.03, SD=.5 7); t(423 )= .638, p=.730 for the participants assigned to the predato r language randomization This shows that parents are showing more of a consensus against sex offenders compared to their non parent participant counterparts, pr ovid ing support for the parent hypothesis (8c) However, the relationship only exists for the offender t test rather than for the predator t test, which does not provide support for the similarity of offender or predator measures (8b) It was hypothesized th at there would be no difference between the two randomly assigned groups. For gender, t he first t test shows that there is no significant difference in consensus between the male (M=2.97, SD=.55 ) and female participants (M=3.17, SD=.57); t(450)= 3.823, p=. 374 for the participants assigned to the offender language randomization There was a weakly significant difference in consensus between male (M=2.93, SD=.54 ) and female participants (M=3. 18, SD=.60); t(424)= 4.383 p=. 024* for the participants assigned t o the predator language randomization This only provides partial support for the gender hypothesis, which stated that women would be more unified than men (8d) 152

PAGE 153

Tab le 4 24 Independent Sample T Test for Parental Status and the Element of Concern Offender Randomization Mean SD N df T Sig. Parents 3.08 .59 250 450 .770 .020* Non Parents 3.04 .53 202 450 .770 Offender Randomization Male Female 2.97 3.17 .55 .57 237 214 450 450 3.823 3.823 .374 Predator Randomization Parents 3.06 .59 228 423 .638 .730 Non Parents 3.03 .57 197 423 .638 Predator Randomization Male Female 2.93 3.18 .54 .60 227 198 423 423 4.383 4.384 .024* *** p<.001 **p < .01 *p < .05 The next step in analyzing this model is to run two OLS regression to predict the moral panic element of consensus. One model will be run for the offender randomization and the other for the predator randomization. The additive registry knowledge variable will still serve as the independent variable and all o f the previously used controls will remai n control variables. Table 4 25 show s both of those OLS Regressions. Overall both models are significant the offender regression is moderately significant at the .01 alpha level and the predator regression is st rongly significant at the .001 alpha level. The predictor varia bles are able to account for only 7.0% (offender) and 11.4 % (predator) of the respective explained variance, as shown by the R Squares. Even though both models are significant, they are the w eakest of the OLS Regression analyses examining the five elements of the moral panic. For the offender model, only the registry knowledge and stereotypical offender variable s are significant predictor s for participant consensus (significant at the .05 and .01 alpha level s respectively ). The lack of significant relationships supports the hypothesis for high levels of consensus for the most part ( 8a ). However, the two hypotheses 153

PAGE 154

exceptions (referencing parent status and gender) were not supported for the o ffender model (8c and 8d) For the predator regression model, three variables proved to be related to consensus; the stereotypical offender (significant at the .001 alpha level), number of school age children (significant at the .05 alpha level), and gen der (significant at the .001 alpha level) There is no significant relationship in the predator model between parental status and the ele ment of consensus so the parent hypothesis which states that parent participants will be more unified than non parent s is not supported (8c) However, the non significant findings might be an indication of consensus across groups rather than within then. These findings provide support for the overall concept of consensus, just not for the individual hypothesis discusse d for this specific model. Gender was a significant predictor but only for the predator regression model. This only provides partial support for the gender hypothesis, which stated that women would be more unified than men (8d) The registry knowledge va riable did not significantly predict consensus for either the offender randomization or the predator one. There was a difference in direction for each of the registry knowledge variables, with the offender randomization showing a positive relationship and with the predator randomization showing a negative one. However, the predator regression model is stronger and able to explain more of the variance overall. Like previous regression models, the stronger predator model suggests that there is something sy stematically different about the terms "sex ual offender" and "sex ual predator," making the latter more salient in the minds of the participants, but that registry knowledge might be a weaker predicting variable than was 154

PAGE 155

originally intended. The stronger p redator model does not provide pr ovides support for the similarity of offender or predator measures (8b) which states that there should be no difference regarding the participants' response to the two randomly assigned measurement groups Table 4 25 OLS Regression Predicting Element of Consensus (Random Assignment). Variable Offender Randomization Predator Randomization Coefficient Std. Error Beta Coefficient Std. Error Beta Registry Knowledge .025* .013 .091 .007 .007 .049 Stereotypical Offender .050 ** .020 .133 .088 *** .021 .239 Parental Status .00 7 .072 .015 .024 .071 .021 Num. School Age Children .025 .036 .039 .075* .038 .107 Gender .004 .006 .034 .209*** .057 .179 Age .019 .013 .080 .009 .014 .036 Race .006 .065 .005 .019 .064 .016 Ethnicity .097 .087 .053 .006 .004 .068 Education Level .032 .058 .028 .021 .062 .018 Marital Status .070 .059 .062 .053 .060 .045 Income Level .031 .037 .044 .055 .044 .068 SES Status .072 .052 .075 .089 .058 .086 Geographic Region .078 056 .066 .003 .058 .002 Population Size .002 .022 .004 .015 .022 .032 Constant 2.716 148 2.503 .1 46 F Statistic 2.284 ** 3.655 *** R Square .070 .114 *** p<.001 **p < .01 *p < .05 Parental Status (dummy coded: Non Parent/Parent), Race (dummy coded: Non White/White), Ethnicity (dummy coded: Non Hispanic/Hispanic), Education Level (dummy coded: No Bachelor's Degree/Bachelor's Degree), Marital Status (dummy coded: Not Married/Not Married), Geog raphic Region (Not from the South/From the South) Model 7: OLS Regression Model Predicting the Moral Panic Element of Disproportionality For Model 7, several hypotheses have been derived concerning the participants' perceptions of the disproportionate res ponse that is being taken against sex offenders in response to their sexual crimes. This disproportionate response would be reflected in 155

PAGE 156

the harsh, stigmatizing laws that prosecute sex crimes. This set of questions used the implemented random assignment of "sexual offender" and "sexual predator" questions to the participants. First, the reaction to sex offenses and offenders will be disproportionate (9a) Then the disproportionate reactions will extend across social categories and groups, and it will be similar for the group responding to "sex ual offender" and that responding to "sex ual predator" survey items (9b) Next, it is predicted that parents will react more disproportionately than will non parents (9c) and women will react more disproportionate ly than will men (9d) In order to test this hypothesis, frequency analysis, bivariate analyses, independent sample t tests and OLS Regressions will be completed. Table 4 26 shows the u nivariate analysis for the moral panic element of disproportionality a gainst sex offenders living in our communities. Five measures were used to measure disproportionality and were included in the implemented random assignment of "sex ual offender" and "sex ual predator." The disproportionality variables were measured just l ike the moral panic element variables, however three of the measures were reversed coded in order to keep all five of the measures on the same logic trajectory. This means that once the measures are reverse coded, higher response values indicate increased support for the severe sanctions against sex offenders. The results show that for the most part, participants are responding with support for more harsh punishments for sex offenders. This is the first indication that participants are in favor of a stri ct and severe legal atmosphere towards sex offenders and sex predators and it also provides support for the first hypothesis, which suggests that participants will react in a disproportionate manner (9a) 156

PAGE 157

Table 4 26 Univariate Analysis for Moral Panic El ement of Disproportionality (Offender and Predator Randomization). Measure (Offender n = 452) Code Frequency Percent Do you feel that the current state of the sex offender registry is too harsh?* 1 = Definitely Yes 2 = Probably Yes 3 = Probably Not 4 = Definitely Not Missing (Predator Assignment) n = 25 n = 100 n = 191 n = 136 n = 425 2.8% 11.4% 21.8% 15.5% 48.5% (51.5%) (100.0%) Do you feel that the sex offender registry laws should be stricter? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) n = 27 n = 143 n = 174 n = 108 n = 425 3.1% 16.3% 19.8% 12.3% 48.5% (51.5%) (100.0%) Do you think that keeping sex offenders on electronic monitoring/GPS tracking for more than 5 years without a break is too severe a punishment?* 1 = Definitely Yes 2 = Probably Yes 3 = Probably Not 4 = Definitely Not Missing (Predator Assignment) n = 50 n = 109 n = 157 n = 136 n = 425 5.7% 12.4% 17.9% 15.5% 48.5% (51.5%) (100.0%) Do you think that sex offenders should report to law enforcement more than the required two times per year? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) n = 20 n = 81 n = 180 n = 171 n = 425 2.3% 9.2% 20.5% 19.5% 48.5% (51.5%) (100.0%) Do you think that the media overreacts in their reporting of sex offenses when they occur in a community?* Measure (Predator n = 425) 1 = Definitely Yes 2 = Probably Yes 3 = Probably Not 4 = Definitely Not Missing (Predator Assignment) Code n = 62 n = 136 n = 140 n = 114 n = 425 Frequency 7.1% 15.4% 16.0% 13.0% 48.5% (51.5%) (100.0%) Percent Do you feel that the current state of the sex offender registry is too harsh?* 1 = Definitely Yes 2 = Probably Yes 3 = Probably Not 4 = Definitely Not Missing (Offender Assignment) n = 38 n = 95 n = 196 n = 96 n = 452 4.3% 10.8% 22.3% 11.1% 51.5% (48.5%) (100.0%) Do you feel that the sex offender registry laws should be stricter? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 32 n = 134 n = 173 n = 86 n = 452 3.6% 15.4% 19.7% 9.8% 51.5% (48.5%) (100.0%) 157

PAGE 158

Table 4 26 Continued Measure (Offender n = 452) Code Frequency Percent Do you think that keeping sex predators on electronic monitoring/GPS tracking for more than 5 years without a break is too severe a punishment?* 1 = Definitely Yes 2 = Probably Yes 3 = Probably Not 4 = Definitely Not Missing (Offender Assignment) n = 52 n = 125 n = 146 n = 102 n = 452 5.9% 14.3% 16.7% 11.6% 51.5% (48.5%) (100.0%) Do you think that sex predators should report to law enforcement more than the required two times per year? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 28 n = 100 n = 173 n = 124 n = 452 3.3% 11.4% 19.7% 14.1% 51.5% (48.5%) (100.0%) Do you think that the media overreacts in their reporting of sex offenses when they occur in a community?* 1 = Definitely Yes 2 = Probably Yes 3 = Probably Not 4 = Definitely Not Missing (Offender Assignment) n = 58 n = 140 n = 133 n = 94 n = 452 6.6% 16.0% 15.2% 10.7% 51.5% (48.5%) (100.0%) *Indicates that the measure was reverse coded. The next step in this analysis was to create two scales from the five offender and five predator measures. The two scales were created by adding the scores of the five measures across participants and then divide that score by five. Once the scales were completed then, two sets of bivariate correlations were completed to look at disproportionality against the control variables. In the first bivariate correlation (Table B 32, Appendix) five control variables show a significant correlation with the Disprop ortionality scale (Offender sub sample ). Parental status ( r = .154, significant at the .001 alpha level), number of school age children ( r = .101, significant at the .05 alpha level), age ( r = .210, significant at the .001 alpha level) education level (r = .124, significant at the .01 alpha level), and population size ( r = .117, significant at the .01 alpha level) all show a significant correlation with disproportionality (offender sub sample ). 158

PAGE 159

The various hypotheses regarding disproportionality are s upported at different levels. First, because five of the variables showed a significant correlation, the hypothesis regarding similarities of disproportionality across all social categories and groups is not supported (9b) Second, the parent hypothesis stating that parents will react more disproportionately compared to non parents is supported (9c) Finally, the gender hypothesis which states that women will react more disproportionately than men is not supported because the correlation was not significant (9d) In the second bivariate correlation (Table B 33, Appendix) five control variables show a significant correlation with the Disproportionality scale (Predator sub sample ). Parental status ( r = .156, significant at the .001 alpha level), number of school age children ( r = .118, significant at the .05 alpha level), gender ( r = .292, significant at the .001 alpha level) and age ( r = .151, significant at the .001 alpha level), marital status ( r = .111, significant at the .05 alpha level) all show a significant correlation with Disproportionality (Predator sub sample ). The various hypotheses regarding disproportionality are supported at different levels. First, because five of the variables showed a significant correlation, the hypothesis r egarding similarities of disproportionality across all social categories and groups is not supported (9b) Second, the parent hypothesis stating that parents will react more disproportionately compared to non parents is supported (9c) Finally, the gender hypothesis which states that women will react more disproportionately than men is supported because the correlation was not significant in the offender model is significant in the predator one (9d) 159

PAGE 160

Next, two independent sample t tests were conducted to see if there is any significant difference between the parent and non parent groups for the disproportionality scales Table 4 27 shows the results of the two t tests. The first shows that there is a strongly significant differ ence in disproportionality between the parents (M=2.98, SD=.66) and non parents (M=2.78, SD=.64); t(450 )= 3.260, p=.001*** for the participants who were assigned to the offender language randomization. For the second t test, there also was a strongly sign ificant difference between parents (M=2.86, SD=.64) and non parents (M=2.65, SD=.65); t(423 )= 3.209, p=.001*** for the predator randomization. This shows that parents are more likely to support strict and severe laws against sex offenders compared to non parents, which provide s support for the parent hypothesis which states that parents will react more disproportionately than non parents (9c) Additionally the results provi de support for the similarity of offender and predator measures suggesting that citizens do no distinguish much between the two (9b) For the gender t test, t he first shows that there is a moderate significant difference in disproportionality between the male (M=2.64, SD=.39 ) and female participants (M=2.43, SD=.65 ); t(450 )=2.803, p =.005 ** for the participants assigned to the offender language randomization. For the second t test, there no significant difference in disproportionality between male (M=2.53, SD=.46 ) and female participants (M=2.46, SD=.40 ); t(423 )=1.827, p=.068 for the participants assigned to the predator language randomization. These findings provide no su pport for the gender hypothesis (9d) The hypothesis predicted that women would react m ore disproportionately than 160

PAGE 161

men. The means for the men were higher in both the offender and predator conditions, but the difference in the offender condition was the only one to show significance. Table 4 27 Independent Sample T Test for Parental Status and the Element of Disproportiona lity. Offender Randomization Mean SD N df T Sig. Parents 2.98 .66 250 450 3.260 .001*** Non Parents 2.78 .64 202 450 3.260 Offender Randomization Males Females 2.54 2.43 .39 .45 237 214 450 450 2.803 2.803 .005** Predator Randomization Parents 2.86 .64 228 42 3 3.209 .001*** Non Parents 2.65 .65 197 423 3.209 Predator Randomization Males Females 2.53 2.46 .46 .40 227 198 423 423 1.827 1.827 .068 *** p<.001 **p < .01 *p < .05 The next step in analyzing this model is to run two OLS regression to predict the moral panic element of disproportionality. One model will be run for the offender randomization and the other for the predator randomization. The additive registry knowledge variable will still serve as the independe nt variable and all of the previously used controls will remain control variables. Table 4 28 shows results for both of those OLS Regressions. Overall both models are strongly significant at the .001 alpha level. The predictor varia b les are able to acco unt for 14.0 % (offender) and 16.3 % (predator) of the respective explained variance, as shown by the R Squares. For the offender model, there were two significant predictors for perceived disproportionality; the stereotypical offender (significant at the 001 alpha level) and age (significant at the .05 alpha level) Due to the positive but non significant relationship found between the parental status variable and the element of disproportionality, the results of this regressi on do not support the parent hypothesis which states that parent s will react more disproportionally than non parents (9c ) Gender is not a significant predictor variable for the offender model, thus not providing support for the gender hypothesis (9d ). 161

PAGE 162

Table 4 28 OLS Regression Predicting Element of Disproportionality (Random Assignment). Variable Offender Randomization Predator Randomization Coefficient Std. Error Beta Coefficient Std. Error Beta Registry Knowledge .005 .0 15 .015 .01 9 .008 .066 Stereotypical Offender .115*** .022 .270 .095*** .022 .225 Parental Status .136 .080 .103 .119 .078 .091 Num. School Age Children .032 .040 .042 .040 .041 .050 Gender .008 .006 .055 .327*** .063 .250 Age .029* .015 .102 .012 .015 .042 Race .090 .072 .066 .090 .070 .069 Ethnicity .043 .097 .021 .006 .004 .060 Education Level .068 .064 .051 .046 .068 .035 Marital Status .049 .066 .038 .084 .065 .064 Income Level .010 .041 .012 .025 .048 .027 SES Status .008 .058 .007 .100 .063 .085 Geographic Region .076 .062 .055 .109 .064 .080 Population Size .042 .024 .080 .016 .024 .032 Constant 2.344 165 2.121 179 F Statistic 4.911 *** 5.518 *** R Square .140 .163 *** p<.001 **p < .01 *p < .05 Parental Status (dummy coded: Non Parent/Parent), Race (dummy coded: Non White/White), Ethnicity (dummy coded: Non Hispanic/Hispanic), Education Level (dummy coded: No Bachelor's Degree/Bachelor's Degree), Marital Status (dummy coded: Not Married/No t Married), Geographic Region (Not from the South/From the South) For the predator regression model, stereotypical offender and gender are both significant predictors for participant perceived disproportionality (at the .001 alpha level) Like in the o ffe nder model, a positive yet non significant relationship between parental status and the element of disproportionality provides no support for the parent hypothesis (9c). Gender was a significant predictor in the predator regression only. This finding onl y provides partial support for the gender hypothesis, which suggested that women would react more disproportionately than men (9d). 162

PAGE 163

The registry knowledge variable did not significantly predict disproportionality for either the offender randomization or th e predator one. There was a difference in direction for each of the registry knowledge variables, with the offender randomization showing a negative relationship and with the predator randomization showing a positive one. However, the predator regression model is stronger and able to explain more of the variance for disproportionality overall. Like previous regression models, the stronger predator model suggests that there is something systematically different about the terms "sex ual offender" and "sex ua l predator," making the latter more salient in the minds of the participants, but that registry knowledge might be a weaker predicting variable than was originally intended. The stronger predator model does not provide pr ovides support for the similarity of offender or predator measure hypothesis (9b). Model 8: OLS Regression Model Predicting the Moral Panic Element of Volatility For Model 8, there are no a priori hypotheses being developed. Instead, this is an exploratory model that tests volatility as a construct. This set of questions used the implemented random assignment of "sexual offender" and "sexual predator" questions to the participants. This is the most unstable of Cohen's elements, and therefore wi ll be addressed in an exploratory form only. In order to test this hypothesis, frequency analysis, bivariate analyses, independent sample t tests and OLS Regressions will be completed. Table 4 29 shows the u nivariate analysis for the moral panic element o f volatility against sex offenders living in our communities. Five measures were used to measure volatility and were included in the implemented random assignment of "sex offender" and "sex predator." The volatility variables were measured just like the other moral panic variables. 163

PAGE 164

Table 4 2 9 Univariate Analysis for Moral Panic Element of Volatility (Offender and Predator Randomization). Measure (Offender n = 452) Code Frequency Percent Do you think that law enforcement reacts quickly when a sexual of fense takes place? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) n = 63 n = 165 n = 181 n = 43 n = 425 7.2% 18.8% 20.6% 4.9% 48.5% (51.5%) (100.0%) Do you think that legislators work fast enough to get necessary registry laws passed to further keep track of sex offenders? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) n = 72 n = 198 n = 146 n = 36 n = 425 8.2% 22.6% 16.6% 4.1% 48.5% (51.5%) (100.0%) Do you think that the media reports sex offense cases too quickly before all of the facts are gathered? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) n = 40 n = 114 n = 204 n = 94 n = 425 4.5% 12.9% 23.1% 11.0% 48.5% (51.5%) (100.0%) Do you think that the quick response of the media makes communities safer because people are made aware of the sex offense? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) n = 45 n = 144 n = 198 n = 65 n = 425 5.1% 16.4% 22.5% 7.5% 48.5% (51.5%) (100.0%) Do you think that police are too slow to catch sex offenders when sex offenses take place? Measure (Predator n = 425) 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Predator Assignment) Code n = 25 n = 148 n = 206 n = 73 n = 425 Frequency 2.9% 16.9% 23.5% 8.3% 48.5% (51.5%) (100.0%) Percent Do you think that law enforcement reacts quickly when a sexual offense takes place? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 35 n = 150 n = 198 n = 42 n = 452 4.0% 17.1% 22.6% 4.8% 51.5% (48.5%) (100.0%) Do you think that legislators work fast enough to get necessary registry laws passed to further keep track of sex predators? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 59 n = 167 n = 173 n = 26 n = 452 6.7% 19.1% 19.7% 3.0% 51.5% (48.5%) (100.0%) 164

PAGE 165

Table 4 29 Continued Measure (Offender n = 452) Code Frequency Percent Do you think that the media reports sex offense cases too quickly before all of the facts are gathered? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 29 n = 101 n = 187 n = 108 n = 452 3.3% 11.5% 21.3% 12.4% 51.5% (48.5%) (100.0%) Do you think that the quick response of the media makes communities safer because people are made aware of the sex offense? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 48 n = 135 n = 187 n = 55 n = 452 5.5% 15.4% 21.3% 6.3% 51.5% (48.5%) (100.0%) Do you think that police are too slow to catch sex offenders when sex offenses take place? 1 = Definitely Not 2 = Probably Not 3 = Probably Yes 4 = Definitely Yes Missing (Offender Assignment) n = 29 n = 156 n = 196 n = 44 n = 452 3.3% 17.8% 22.4% 5.0% 51.5% (48.5%) (100.0%) The results show that for the most part, participants are responding in somewhere in the middle with only a slight inclination to respond positively meaning the answers are more frequently in the "Probably No" and "Probably Yes" categories rather than th e more definitive options. The next step in this analysis was to create two scales from the five offender and five predator measures. The two scales were created by adding the scores of the five measures across participants and then divide that score by five. Once the scales were completed then two bivariate correlations were completed in order to determine whether any of the control variables are correlated with the Volatility scale. Since there was a random assignment of "sexual offender" and "sexual predator" implemented for the volatility measures, the bivariate correlations were conducted with that random assignment in mind In the first set of bivariate correlations (Table B 34, Appendix) only two control variables show a significant 165

PAGE 166

correlation w ith Volatility (Offender sub sample ). Age (r = .094, significant at the .05 alpha level) and education level ( r = .100, both significant at the .05 alpha level) have a significant correlation with volatility. There were no a priori hypotheses for volati lity. In the second set of bivariate correlations (Table B 35, Appendix) there were no significant correlations between the control variables and Volatility (Predator sub sample ). Next, two independent sample t tests were conducted to see if there is any significant difference between the parent and non parent groups for these scaled variables. Table 4 3 0 shows the results of the two t tests. The first shows that there is no significant difference between the parents (M=2.60, SD=.40) and non parents (M=2 .57, SD=.43); t(450 )= .918, p=.147 for the offender randomization. There was also no significant difference between parents (M=2.61, SD=.42) and non parents (M=2.61, SD=.40); t(423 )=.040, p=.645 for the predator randomization. The means of each group are similar and t his suggests that parents and non parents equally feel that the criminal justice system is not reacting quickly enough in response to sexual offenses. Although there is no significant different between the two participant groups, this still provides support to the overall project objective that predicted a moral panic is sustained among community members in regard to sex offenders living in communities. The second set of t tests shows that there is a moderate significant difference in volat ility between male (M=2.64, SD=.41 ) and female participants (M=2.53, SD=.40); t(450)= 2.823, p=.005** for the offender randomization. There was no significant difference between male (M=2.64, SD=.40 ) and female participants (M=2.58, SD=.41); t(423)=1.457 p=. 146 for the predator randomization. In these two t test s the results suggest that males are more likely than females to exert volatile reactions to sex crimes. 166

PAGE 167

This was not a predicted relationship, but it does suggest that a gender component is imb edded in the data, like in the previous models this time however it is in the reverse direction than what was previously discussed. The significant t test between genders for the offender randomization is also a bit different from that which is normally seen. Typically the predator model is the stronger of the two but in this instance, the offender was stronger. Table 4 3 0 Independent Sample T Test for Parental Status and the Element of Volatility Offender Randomization Mean SD N df T Sig. Parents 2.60 .40 250 450 .918 .147 Non Parents 2.57 .43 202 450 .918 Offender Randomization Male Female 2.64 2.53 .41 .40 237 214 450 450 2.823 2.834 .005** Predator Randomization Parents 2.61 .42 228 423 .040 .645 Non Parents 2.61 .40 197 423 .040 Predator Randomization Male Female 2.64 2.58 .40 .41 227 198 423 423 1.457 1.457 .146 *** p<.001 **p < .01 *p < .05 The next step in analyzing this model is to run two OLS regression to predict the moral panic element of volatility. One model will be run for the offender randomization and the other for the predator randomization. The additive registry knowledge variable will still serve as the independent variable and all of the previously used controls will remain control variables Table 4 3 1 shows both of those OLS Regressions. Overall both models are significant at the .0 0 1 alpha level (offender) and the .05 alpha level (predator) respectively However, based on the R Squares for the first time, the offender model is s tronger than the predator model but both are relatively low. The predictor variables are able to account for 8.7 % (o ffender) and 5.7 % (predator) of the respective explained variance, as shown by the R Squares. For the offender model, registry knowledge (signifi cant at the .05 alpha level), the stereotypical offender variable 167

PAGE 168

and ethnicity (both significant at the .01 alpha level) are significant predictors for the element of volatility. The parental status variable is not significant for the model this finding is consistent with the results of the t tests, which also showed no significant difference between the two participant groups for the offender randomization. For the predator regression model, the stereotypical sex offender (significant at the .05 level) i s the only significant predictor variable for the element of volatility. There is a negative relationship between the stereotypical sex offender variable and volatility, which indicates the more in accurate the perceived offender profile becomes, the more likely the participants are to feel that the criminal justice system is not reacting quickly enough when a sexual offense occurs. The registry knowledge variable did significantly predict volatility for the offender model but did not significant ly predic t it for the predator model. There was a difference in direction for each of the registry knowledge variables, with the offender randomization showing a positive relationship and with the predator randomization showing a negative one. However, the predat or regression model is weaker than the offender model and able to explain less of the variance overall. Unlike previous regression models where the predator model was the stronger of the two, this model does not suggest that there is a robust difference a bout the terms "sex ual offender" and "sex ual predator" and how participants perceive them. Gender was not a significant predictor in either regression model for the prediction of volatility. This is also a reversal compared to the other analyses conducte d on the five elements of Cohen's moral panic. In the last four regressions, gender was a significant predictor for either the offender model, the regression model, or for both. In this instance, it was not significant in either instance. 168

PAGE 169

Table 4 3 1 OLS Regression Predicting Element of Volatility (Random Assignment). Variable Offender Randomization Predator Randomization Coefficient Std. Error Beta Coefficient Std. Error Beta Registry Knowledge .020* .0 10 .100 .015 .010 .075 Stereotypical Offender .040 ** .014 .140 .024 .015 111 Parental Status .064 .051 .078 .001 .051 .001 Num. School Age Children .012 .026 .025 .003 .027 .007 Gender .005 .004 .063 .027 .041 .033 Age .014 .009 .083 .001 .010 .006 Race .079 .046 .093 .098* .046 .119 Ethnicity .156** .063 .119 .001 .003 .017 Education Level .073 .042 .088 .016 .045 .019 Marital Status .005 .042 .006 .018 .043 .022 Income Level .008 .027 .017 .031 .032 .053 SES Status .026 .037 .037 .010 .042 .014 Geographic Region .018 .040 .022 .020 .042 .024 Population Size .019 .016 .057 .013 .016 .041 Constant 2.934 .184 2.470 .198 F Statistic 2.894* ** 2.229* R Square .087 .0 57 *** p<.001 **p < .01 *p < .05 Parental Status (dummy coded: Non Parent/Parent), Race (dummy coded: Non White/White), Ethnicity (dummy coded: Non Hispanic/Hispanic), Education Level (dummy coded: No Bachelor's Degree/Bachelor's Degree), Marital Status (dummy coded: Not Married/Not Marri ed), Geographic Region (Not from the South/From the South) Model 9: OLS Regression Model Predicting the Community Attitudes Toward Sex Offenders (CATSO) Scale For Model 9, several hypotheses have been derived concerning the participants' perceptions of and attitudes toward sex offenders. First, citizens will have highly negative attitudes for the overall scale and on each of the four sub dimensions (social isolation, capacity for change, severity /dangerousness, and deviancy) (11a) Then, it is predicted that these attitudes will be similar across social categories and groups (11b) Parents will hold more negative attitudes on the CATSO scale and sub dimensions than will nonparents (11c) Women w ill hold more negative attitudes on the CATSO scale 169

PAGE 170

and sub dimensions than will men (11d) These measures were not presented with the sexual offender or sexual predator random assignment. In order to test this hypothesis, frequency analyses, bivariate a nalyses, independent sample t tests and OLS Regressions will be completed. Table 4 3 2 shows the u nivariate analysis for the CATSO Scale. Eighteen measures were used to measure participant's attitudes toward and perceptions of sex offenders. The CATSO Sca le variables were measured using a six point Likert scale, just as the original Church et al. scale was measured. Response options include "Strongly Disagree," "Disagree," "Probably Disagree," "Probably Agree," "Agree," and "Strongly Agree." Like the fiv e elements of Cohen's moral panic, participants are responding in somewhere in the middle of the response options. With only a few exceptions, participants are largely split in how they feel about sex offenders. Some of the more controversial measures el icited a strong response from participants these measures were largely those that were reverse coded as originally done by Church et al. (2008). For example, the statement "Only a few sex offenders are dangerous," showed strong disagreement among particip ants. This indicates that most people believe the opposite to be true about sexual offenders that most sex offenders are dangerous, no matter the type of offense that is committed. This blanket statements and others show the misconceptions associated w ith community perceptions of sex offenders and of the registry itself. Overall, a definitive conclusion about the first hypothesis which stated that citizens would have highly negative attitudes for the overall scale cannot be made at this time (11a) Further analysis needs to be conducted for a more certain conclusion can be made. 170

PAGE 171

Table 4 3 2 Univariate A nalysis for the Community Attitudes Toward Sex Offenders (CATSO) Scale. Measure (n = 877) Code Frequency Percent Social Isolation (6) Sex offenders prefer to stay home alone rather than to be around lots of people. 1 = Strongly Disagree 2 = Disagree 3 = Probably Disagree 4 = Probably Agree 5 = Agree 6 = Strongly Agree n = 42 n = 99 n = 290 n = 264 n = 131 n = 51 4.8% 11.3% 33.1% 20.1% 14.9% 5.8% (7) Most sex offenders do not have close friends. 1 = Strongly Disagree 2 = Disagree 3 = Probably Disagree 4 = Probably Agree 5 = Agree 6 = Strongly Agree n = 60 n = 170 n = 295 n = 198 n = 122 n = 32 6.8% 19.4% 33.6% 22.6% 13.9% 3.6% (8) Sex offenders have difficulty making friends even if they try real hard. 1 = Strongly Disagree 2 = Disagree 3 = Probably Disagree 4 = Probably Agree 5 = Agree 6 = Strongly Agree n = 47 n = 157 n = 285 n = 228 n = 123 n = 37 5.4% 17.9% 32.5% 26.0% 14.0% 4.2% (14) Most sex offenders are unmarried men. 1 = Strongly Disagree 2 = Disagree 3 = Probably Disagree 4 = Probably Agree 5 = Agree 6 = Strongly Agree n = 86 n = 189 n = 283 n = 207 n = 85 n = 27 9.8% 21.6% 32.3% 23.6% 9.7% 3.1% (16) Most sex offenders keep to themselves. 1 = Strongly Disagree 2 = Disagree 3 = Probably Disagree 4 = Probably Agree 5 = Agree 6 = Strongly Agree n = 52 n = 128 n = 292 n = 248 n = 128 n = 29 5.9% 14.6% 33.3% 28.3% 14.6% 3.3% Capacity to Change (1) With support and therapy, someone who committed a sexual offense can learn to change their behavior.* 1 = Strongly Agree 2 = Agree 3 = Probably Agree 4 = Probably Disagree 5 = Disagree 6 = Strongly Disagree n = 54 n = 88 n = 175 n = 286 n = 198 n = 76 6.2% 10.0% 20.0% 32.6% 22.6% 8.7% (2) People who commit sex offenses should lose their civil rights (e.g. voting and privacy). 1 = Strongly Disagree 2 = Disagree 3 = Probably Disagree 4 = Probably Agree 5 = Agree 6 = Strongly Agree n = 83 n = 129 n = 232 n = 198 n = 142 n = 93 9.5% 1 4.7% 26.5% 22.6% 16.2% 10.6% (11) Trying to rehabilitate sex offenders is a waste of time. 1 = Strongly Disagree 2 = Disagree 3 = Probably Disagree 4 = Probably Agree 5 = Agree 6 = Strongly Agree n = 95 n = 146 n = 286 n = 175 n = 110 n = 65 10.8% 16.6% 32.6% 20.0% 12.5% 7.4% 171

PAGE 172

Table 4 3 2 Continued Measure (n = 877) Code Frequency Percent (12) Sex offenders should wear tracking devices so that their location can be pinpointed at any time. 1 = Strongly Disagree 2 = Disagree 3 = Probably Disagree 4 = Probably Agree 5 = Agree 6 = Strongly Agree n = 64 n = 66 n = 186 n = 235 n = 177 n = 149 7.3% 7.5% 21.2% 26.8% 20.2% 17.0% (18) Convicted sex offenders should never be released from prison. 1 = Strongly Disagree 2 = Disagree 3 = Probably Disagree 4 = Probably Agree 5 = Agree 6 = Strongly Agree n = 109 n = 144 n = 288 n = 153 n = 99 n = 84 12.4% 16.4% 32.8% 17.4% 11.3% 9.6% Severity/Dangerousness (4) Male sex offenders should be punished more severely then female sex offenders.* 1 = Strongly Agree 2 = Agree 3 = Probably Agree 4 = Probably Disagree 5 = Disagree 6 = Strongly Disagree n = 243 n = 222 n = 176 n = 118 n = 86 n = 32 27.7% 25.3% 20.1% 13.5% 9.8% 3.6% (9) The prison sentences sex offenders receive are much too long when compared to the sent ence lengths for other crimes.* 1 = Strongly Agree 2 = Agree 3 = Probably Agree 4 = Probably Disagree 5 = Disagree 6 = Strongly Disagree n = 162 n = 186 n = 271 n = 156 n = 75 n = 27 18.5% 21.2% 30.9% 17.8% 8.6% 3.1% (13) Only a few sex offenders are dangerous.* 1 = Strongly Agree 2 = Agree 3 = Probably Agree 4 = Probably Disagree 5 = Disagree 6 = Strongly Disagree n = 44 n = 118 n = 191 n = 188 n = 191 n = 145 5.0% 13.5% 21.8% 21.4% 21.8% 16.5% (15) Someone who uses emotional control when committing a sex offense is not as bad as someone who uses physical control when committing a sex offense.* 1 = Strongly Agree 2 = Agree 3 = Probably Agree 4 = Probably Disagree 5 = Disagree 6 = Strongly Disagree n = 220 n = 185 n = 198 n = 164 n = 79 n = 31 25.1% 21.1% 22.6% 18.7% 9.0% 3.5% (17) A sex offense committed against someone the perpetrator knows is less serious than a sex offense committed against a stranger.* 1 = Strongly Agree 2 = Agree 3 = Probably Agree 4 = Probably Disagree 5 = Disagree 6 = Strongly Disagree n = 363 n = 161 n = 137 n = 124 n = 70 n = 22 41.4% 18.4% 15.6% 14.1% 8.0% 2.5% Deviancy (3) People who commit sex offenses want to have sex more often then the average person. 1 = Strongly Disagree 2 = Disagree 3 = Probably Disagree 4 = Probably Agree 5 = Agree 6 = Strongly Agree n = 84 n = 171 n = 237 n = 220 n = 115 n = 50 9.6% 19.5% 27.0% 25.1% 13.1% 5.7% 172

PAGE 173

Table 4 3 2 Continued. Measure (n = 877) Code Frequency Percent (5) A lot of sex offenders use their victims to create pornography. 1 = Strongly Disagree 2 = Disagree 3 = Probably Disagree 4 = Probably Agree 5 = Agree 6 = Strongly Agree n = 33 n = 75 n = 245 n = 282 n = 169 n = 73 3.8% 8.6% 27.9% 32.2% 19.3% 8.3% (10) Sex offenders have high rates of sexual activity. 1 = Strongly Disagree 2 = Disagree 3 = Probably Disagree 4 = Probably Agree 5 = Agree 6 = Strongly Agree n = 28 n = 96 n = 288 n = 291 n = 132 n = 42 3.2% 10.9% 32.8% 33.2% 15.1% 4.8% *Indicates that the measure was reversed coded, as per the original Church et al. scale. The next step in this analysis was to create four scales from the eighteen measures. In replicating the scales used by Church et al. (2008), the eighteen items were broken down into four constructs Social Isolation (includes items 6, 7, 8, 14 and 16), Capacity to Change (items 1*, 2, 11, 12, and 18), Severity/Dangerousness (4*, 9*, 13*, 15* and 17*), and Deviancy (items 3, 5 and 10). These four scales were created by adding the number of measures present for each construct and then by di viding by the number of measures. Each of the scales is analyzed independently of one another to examine each of the constructs, but ultimately they are recombined in an index to examine the totality of negative attitudes toward sex offenders. In order t o examine the combined attitudes, the four scales were added together to create a Total Index of Negative Attitudes variable From this recombination the results show a total score, where the higher the score indicates more negative attitudes. Once the sc ales were completed then five bivariate correlations were run to test the control variables against the four constructs and the total index of negative aptitudes. In the first bivariate correlation (Table B 36, Appendix), three variables showed a signific ant correlation with the first CATSO construct, Social Isolation. Age ( r 173

PAGE 174

= .157, significant at the .001 alpha level), educational level (r = .088, significant at the .01 alpha level) and marital status ( r = .084, significant at the .01 alpha level) all provided a significant correlation with Social Isolation. In the second bivariate correlation (Table B 37, Appendix) three control variables show a significant correlation with the CATSO construct, Capacity to Ch ange Parental Status ( r = .176, significant at the .01 alpha level), educational level ( r = .066, significant at the .05 alpha level), marital status ( r = .125, significant at the .001 alpha level) all provide a significant correlation between themselves and the second CATSO construct of Capacity to Change. In the third bivariate correlation (Table B 38, Appendix) four control variables show significant correlations with the CATSO construct, Severity/Dangerousness Parental status ( r = .080, significan t at the .05 alpha level), age (r = .200, significant at the .001 alpha level), education level ( r = .150, significant at the .001 alpha level), and socio economic status ( r = .081, significant at the .05 alpha level) are all significantly correlated with the third CATSO construct of Severity/Dangerousness. In the fourth bivariate correlation (Table B 39, Appendix) two control variables show significance with the CATSO construct of Deviancy Parental status (r = .106, significant at the .01 alpha level ) and number of school age children ( r = .088, significant at the .01 alpha level) both provide a significant correlation to the fourth construct of Deviancy. In the fifth bivariate corr elation (Table B 40, Appendix), four control variables are significant in the overall Total Index of Negative Attitudes Parenta l status (r = .125) age (r = .126) education level (r = .116) and marital status (r = .107) ( all significant at 174

PAGE 175

the .001 alpha level), are all significantly correlated with the Total Index of Ne gative Attitudes. The various hypotheses regarding the CATSO scale and its sub constructs are supported at different levels. First, because all five bivariate correlations had two or more significant variables, the hypothesis regarding similarities of pa rticipant responses across all social categories and groups is only partially supported (11b) Second, the parent hypothesis stating that parents will have more negative attitudes toward sex offenders compared to non parents is supported due to the si gnificance of parental status in four of the five bivariate correlations (11c) Finally, the gender hypothesis which states that women will have more negative attitudes toward sex offenders than men is not supported due to the absence of a significant gender correlation for any of the five correlations (11d) Next, five independent sample t tests were conducted to see if there is any significant difference between the parent and non parent groups for the four CATSO construct scaled variables and for the total index of negative attitudes toward sex offenders. Table 4 3 3 shows the results of the five t tests. The first shows that there is no significant difference between the parents (M=3.37, SD=.96) and non parents (M=3.32, SD=.92); t(875)= .854, p=. 224 for the Social Isolation construct. This suggests that both participants groups are equal in their beliefs that sex offenders are introverted, isolated individuals with no social life. In the second t test, there was a strongly significant differenc e between parents (M=3.71, SD=.83) and non parents (M=3.40, SD=.88); t(875)= 5.280, p=.000*** for the Capacity to Change construct. This indicates that parents are more likely than non parents to believe that sex offenders are incapable of change. 175

PAGE 176

In th e third t test, there was a weakly significant difference between parents (M=2.96, SD=.84) and non parents (M=2.83, SD=.81); t(875)= 2.367, p=.018 for the Severity/Dangerousness construct. This indicates that parents are more likely to believe sex offend ers to be dangerous individuals who commit serious crimes, in comparison to non parents. The fourth t test shows a moderately significant difference between parents (M=3.66, SD=.96) and non parents (M=3.45, SD=.97); t(875)= 3.145, p=.002** for the Devian cy construct. This suggests that parents believe sex offenders to exhibit high levels of sexually deviancy in general, in comparison to non parents. Finally in the Total Index of Negative Attitudes, there was also a strongly significant difference between parents (M=13.71, SD=2.77) and non parents (M=13.01, SD=2.77); t(875)= 3.715, p=.000*** Overall, this suggests that parents have more strongly negative attitudes towards sex offenders in comparison to non parents. All but one of the t test shows a significant difference between the parent and non parent participants. These results suggest that parents have more negative attitudes (and have higher levels of support for the four sub constructs) towards sex offenders (11c) This provides very stro ng support for the parent hypothesis which states that parents will more negative attitudes on the CATSO scale and sub dimensions than will non parents. Finally, the gender component of this model needs to be analyzed. The same t tests were run comparing male and female participants. The first shows that there is a strongly significant difference between male (M=3.45, SD=.91 ) and female participants 176

PAGE 177

(M=3.24, SD=.97); t(875)=3.448, p=.001*** for the Social Isolation construct. This indicates that men per ceive sex offenders to be more socially isolated than females. I n the second t test, there was no significant difference betwe en male (M=3.52, SD=.86) and female participants (M=3.63, SD=.87 ); t(875)= 1.823, p=.069 for the Capacity to Change construct. Al though men are less likely to believe that sex offenders are capable of change compared women, the results were not statistically significant. In the third t test, there was a strongly significant difference between male (M=3.08, SD=.83 ) and female participants (M=2.71, SD=.79); t(875)=6.820, p=.001** for the Severity/Dangerousness construct. Males perceive sex offenders to be more severe and dangerous than women. The fourth t test shows no significant difference between male (M=3.59, SD=.98 ) and f emale participants (M=3. 5 4 SD=.97); t(875)=.683, p=.495 for the Deviancy construct. There were no significant conclusions regarding how deviant men and women perceive sex offenders to be. Finally in the Total Index of Negative Attitudes, there was also a moderately significant difference between male (M= 13.64, SD=2.81 ) and female participants (M= 1 3. 11 SD=2.76); t(875)=.2.814, p=.00 5 **. Overall, men have more negative attitudes toward sex offenders than women. Of the four gender t tests on sub dimension two of the sub dimensions (Social Isolation and Severity/Dangerousness) showed a significant difference between genders. However despite the significant differences shown in these t tests, the direction of the tests go in the opposite direction of what w as anticipated. Men scores 177

PAGE 178

were more extreme compared to women across all five measures; therefore the gender hypothesis for this model is not supported (11d) Table 4 3 3 Independent Sample T Tests for Parental Status and the Four Constructs of the CATSO Scale Social Isolation Mean SD N df T Sig. Parents 3.37 .96 479 875 .854 .224 Non Parents 3.32 .92 398 875 .854 Social Isolation Male Female 3.45 3.24 .91 .97 465 412 875 875 3.448 3.448 .001*** Capacity to Change Parents 3.71 .83 479 875 5.280 .000*** Non Parents 3.40 .88 398 875 5.280 Capacity to Change Male Female 3.52 3.63 .86 .87 465 412 875 875 1.823 1.823 .069 Severity/Dangerousness Parents 2.97 .84 479 875 2.367 .018* Non Parents 2.83 .81 398 875 2.267 Severity/Dangerousness Male Female 3.08 2.71 .83 .79 465 412 875 875 6.820 6.820 .000*** Deviancy Parents 3.66 .96 479 875 3.145 .002** Non Parents 3.45 .97 398 875 3.145 Deviancy Male Female 3.59 3.54 .98 .96 465 412 875 875 .682 .682 .495 Total Index of Negative Attitudes Parents 13.71 2.77 479 875 3.715 .000*** Non Parents 13.01 2.77 398 875 3.715 Total Index of Negative Attitudes Male Female 13.64 13.11 2.81 2.76 465 412 875 875 2.814 2.814 .005** *** p <.001 **p < .01 *p < .05 The next step in analyzing this model is to run five OLS regression to predict the four constructs of the CATSO Scale and the total index of negative attitudes toward sex offenders. The additive registry knowledge variable will st ill serve as the independent variable and all of the previously used controls will remain control variables. Table 4 3 4 shows the first two OLS Regressions for Social Isolation and Capacity to Change. Overall both models are strongly significant at the .001 alpha level. The predictor 178

PAGE 179

variab les are able to account for 12.2 % (Social Isolation model) and 12.9 % (Capacity to Change model) of the respective explained variance, as shown by the R Squares. For the Social Isolation model, the stereotypical offender (significant at the .001 alpha level), age (significant at the .01 alpha level), race (significant at the .001 alpha level), and marital status (significant at the .05 alpha level) are all significant predictor variables. While it is not significant, t here is a negative relationship between registry knowledge and Social Isolation, which indicates that as the level knowledge decreases the more likely they are to believe that sex offenders are social isolates who keep to themselves. The parental status variable is not significant for the model this finding is consistent with the results of the Social Isolation t test which also showed no significant difference between the parent and non parents (11c) For the Capacity to Change regr ession model, parental status (significant at the .01 alpha level), gender (significant at the .05 alpha level) and race (significant at the .001 alpha level) are all significant predictor variables. Although it is not significant, t here is a negative rel ationship between registry k nowledge and Capacity to Change, which indicates the as knowledge decreases, the more likely participants are to believe that sex offenders cannot change their sexually deviant behavior. Furthermore, the significant relationshi p between parenta l status and Capacity to Change indicates that parents are more likely to believe that sex offenders cannot change their behavior, compared to non parents. This contradicts the findings of the Capacity to Change t test, which suggested no significant difference b etween parents and non parents (11c) The inclusion of additional control variables in the regression might isolate this 179

PAGE 180

relationship, causing it to emerge. These two OLS Regression models provide initial, but partial, support fo r the parent hypothesis for the overall CATSO analyses. Gender is only a significant predictor for the Capacity to Change Regression model The variable is significant but in a negative direction, which is consistent with the t test for this construct. T his suggests that men are more likely to believe sex offenders as being incapable of change, compared to women. This is contradictory to the gender hypothesis and does not support it (11d) Table 4 3 4 OLS Regression Predicting Social Isolation and Capaci ty to Change Constructs (CATSO Scale). Variable Social Isolation Capacity to Change Coefficient Std. Error Beta Coefficient Std. Error Beta Registry Knowledge .024 .014 .055 .009 .013 .023 Stereotypical Offender .071*** .022 .116 .024 .020 .043 Parental Status .088 .078 .047 212 ** .072 .122 Num. School Age Children .052 .041 .047 .034 .037 .033 Gender .001 .009 .005 .017** .008 .066 Age .038 * .015 .089 .018 .013 .049 Race .339*** .070 .177 .446** .064 .253 Ethnicity .008 .006 .039 .001 .006 .005 Education Level .070 .065 .037 .042 .060 .024 Marital Status .134* .065 .071 .110 .060 .063 Income Level .049 .044 .040 .014 .040 013 SES Status .009 .060 .006 .022 .055 .015 Geographic Region .102 .063 .052 .046 .057 .026 Population Size .002 .024 .003 .034 .022 .050 Constant 4.033 171 3.070 .230 F Statistic 8.550 *** 9.065 *** R Square .122 .129 *** p<.001 **p < .01 *p < .05 Parental Status (dummy coded: Non Parent/Parent), Race (dummy coded: Non White/White), Ethnicity (dummy coded: Non Hispanic/Hispanic), Education Level (dummy coded: No Bachelor's Degree/Bachelor's Degree), Marital Status (dummy coded: Not Married/Not Married), Geographic Region (Not fr om the South/From the South) 180

PAGE 181

Table 4 35 shows the second two OLS Regressions for Severity/Dangerousness and Deviancy. Overall both models are significant at the .001 alpha level. The predictor variab les are able to account for 30.5 % (Severity/Dangerousne ss) and 12.4 % (Deviancy) of the respective explained variance, as shown by the R Squares. For the Severity/Dangerousness model, registry k nowledge, stereotypical offender (both significant at the .001 alpha level), parental status, age (both significant at the .05 alpha level), and race (significant at the .001 alpha level) are significant predictors variables. The parental status shows a significant, positive relationship with Severity/Dangerousness, which suggests that parents are more likely to believ e that sex offenders commit severe crimes and that they are overall a dangerous group. This finding is not consistent with the results of the third t test, which showed no significant difference between the two participant groups for this construct (11c) For the Deviancy regressio n model, stereotypical offender and race ( both significant at the .001 alpha level) are the only significant predictor variables for the model Although it is not significant, t here is a negative relationship between registry kn owledge a nd Deviancy, which indicates that as the knowledge level decreases, the more likely participants are to feel that sex offenders engage in a large amount of sexual deviancy. The parental status does not show a significant, positive relationship with Deviancy, which suggests that parents are no more likely than the non parent group to believe that sex offenders engage in a large amount of sexual deviancy (11c) This is consistent to the results of the fourth t test, which suggested that there was no significant difference between the two groups in terms of their perceptions of Deviancy. 181

PAGE 182

Like the pre vious two OLS Regression models, the findings for these two models provid e partial, support for the hypothesis, which suggests just that participants will have largely negative attitudes toward sex offenders (11a) The results showing parental status as b eing significant for the Severity/D angerousness model only provide limited support for the parent hypothesis (11c) Furthermore, the gender variables are in a negative direction, suggesting that men are more cognizant of the issues measured by the CATSO s cale. The lack of significance attributed to gender within the two models does not support the gender h ypothesis for the CATSO scale (11d) Table 4 35 OLS Regression Predicting OLS Regression Predicting Severity/Dangerousness and Deviancy Constructs (CAT SO Scale). Variable Severity/Dangerousness Deviancy Coefficient Std. Error Beta Coefficient Std. Error Beta Registry Knowledge .036 *** .011 093 .008 .015 .01 7 Stereotypical Offender .1 60 *** .017 .296 .088 *** .023 .139 Parental Status .148 .061 .089 .064 .080 .033 Num. School Age Children .001 .032 .001 .075 .042 .065 Gender .011 .007 .047 .001 .009 .005 Age .029* .012 .083 .011 .015 .027 Race .425*** .054 .251 .489 *** .072 .248 Ethnicity .002 .005 .011 .011 .006 .004 Education Level .105* .051 .062 .032 .067 .016 Marital Status .033 .051 .020 .045 .067 .023 Income Level .055 .034 .051 .014 .045 .011 SES Status .038 .047 .026 .012 .061 .007 Geographic Region .020 .049 .012 .055 .064 .027 Population Size .003 .019 .004 .026 .025 .034 Constant 3.000 .196 4.234 .176 F Statistic 26 927 *** 8.713 *** R Square .305 .124 *** p<.001 **p < .01 *p < .05 Parental Status (dummy coded: Non Parent/Parent), Race (dummy coded: Non White/White), Ethnicity (dummy coded: Non Hispanic/Hispanic), Education Level (dummy coded: No Bachelor's Degree/Bachelor's Degree), Marital Status (dummy coded: Not Married/Not Married), Geographic Region (Not from the South/From the South) 182

PAGE 183

Table 4 3 6 shows the final OLS Regression for the Total Index of Negative Attitudes toward sex offenders. This last regression model is significant at the .001 alpha level. The predictor vari ables are able to account for 23.9 % of the explained variance, as shown by the R Square. Registry Knowledge (significant at the .05 alpha level) stereotypical offender (both significant at the .001 alpha level), parenta l status (significant at the .01 alpha level) and race (significant at the .001 alpha level), are all signi ficant predictors for Total Index of Negative Attitudes. There is a negative relationship between registry knowledge and the Total Index of Negative Attitudes, which indicates that as knowledge decreases, the more negative their attitudes are toward sex offenders become. This provides support for the overall general hypothesis, which predicted that negative attitudes toward sex offenders would be high (11a) Additionally there is a negative relationship between the stereotypical offender variable and the Total Index of Negative Attitudes, which suggests that the less accurate participants are in correctly identifying the stereotypical sex offender, the more negative thei r attitudes are towards sex offenders. The parental status shows a significant, positive relationship with the Total Index of Negative Attitudes, which suggests that parents are more likely than the non parent group to have negative toward sex offenders i n general. This provides partial support to the parent hypothesis (11c) Yet, t his is contradictory to the results of the fifth t test, which suggested that there was no significant difference between the two groups in terms of the Total Index. This dif ference might be due to the addition of control variables in the regression analysis. The gender variable is not significant, but still produces a negative relationship, which indicates males have more negative attitudes toward sex offender 183

PAGE 184

than women do. This is contradictory to the gender hypothesis and does not support it (11d) Since the Total Index of Negative Attitudes is comprised of the four other scales added together, the model reflects a total score for participants across all the measures. U nlike the previous four OLS Regression models, the findings for this final model provide strong support for the overall hypothesis which suggests that participants will have largely negative attitudes toward sex offenders (11a) Table 4 36 OLS Regression Predicting Total Index of Negative Attitudes (CATSO Scale). Variable Total Index of Negative Attitudes Coefficient Std. Error Beta Registry Knowledge .077* .039 .059 Stereotypical Offender .34 9*** .061 .192 Parental Status .550 .216 .098 Num. School Age Children .04 7 .113 .014 Gender .028 .024 .035 Age .031 .040 .027 Race 1.783 *** .193 .313 Ethnicity .006 .017 .010 Education Level .248 .179 .044 Marital Status .322 .178 .058 Income Level .131 .120 .036 SES Status .058 .164 .012 Geographic Region .184 .172 .032 Population Size .055 .066 .025 Constant 16.087 .474 F Statistic 19.251 *** R Square .239 *** p<.001 **p < .01 *p < .05 Parental Status (dummy coded: Non Parent/Parent), Race (dummy coded: Non White/White), Ethnicity (dummy coded: Non Hispanic/Hispanic), Education Level (dummy coded: No Bachelor's Degree/Bachelor's Degree), Marital Status (dummy coded: Not Married/Not Married), Geographic Region (Not from the South/From the South) Model 10: OLS Regression Model Predicting Registry Support For Model 10, several hypotheses have been derived concerning the participants' support for the sex offender registry. This set of questions used the implemented random assignment of "sexual offender" and "sexual predator" questions 184

PAGE 185

to the participants. For this model, it is hypothesized that support for the registry will be high (12a) More specifically, the support will be similar across social categories and groups, and it will be similar for both the group responding to "sex ual offenders" and that responding to "sex ual predators (12b). The exceptions to the similarity are that p arents will be more supportive of th e registry than will nonparents (12c) and w omen will be more supportive of the registry than will men (12d) In order to test these two hypotheses, u nivariate analysis bivariate analysis, independent sample t tests and OLS Regressions will be completed. Table 4 37 shows the u nivariate analysis for the registry support measures. Twelve measures were used to measure registry support and were included in the implemented random assignment of "sex offender" and "sex predator." The registry support variables were measured using a five point Likert Scale which included the response options o f "Strongly Disagree," "Disagree," "Unsure," "Agree," and "Strongly Agree". The results show that for the most part, participants are providing "Unsure" responses. Dependent on the question, there is some indication that the participants are providing mo re positive responses meaning the answers are more frequently in the "Agree" or "Strongly Agree" categories rather than in the "Disagree" or "Strongly Disagree" categories. This can be seen in some of the stronger measures such as the children are safer measure, or the residency restrictions measure. These frequency analyses are the first indication that participants support the sex offender registry system in their state. However, because a lot of participants were unsure in their responses, this does not provide support for the general hypothesis which stated that registry support would be high (12a) 185

PAGE 186

Table 4 37 Univariate A nalysis for Registry Support (Offender and Predator Randomization). Measure (Offender n = 452) Code Frequency Percent R eforms should be made to the sex offender registry. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Predator Assignment) n = 6 n = 40 n = 137 n = 197 n = 72 n = 425 0.7% 4.6% 15.5% 22.5% 8.2% 48.5% (51.5%) (100.0%) The sex offender registry is effective in reducing sex offender reoffending. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Predator Assignment) n = 32 n = 95 n = 140 n = 151 n = 34 n = 425 3.6% 10.8% 16.0% 17.2% 3.9% 48.5% (51.5%) (100.0%) The sex offender registry makes life very difficult for sex offenders living in the community. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Predator Assignment) n = 29 n = 76 n = 99 n = 196 n = 52 n = 425 3.3% 8.7% 11.3% 22.3% 5.9% 48.5% (51.5%) (100.0%) Children are safer if the locations of sex offenders are known. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Predator Assignment) n = 17 n = 50 n = 92 n = 175 n = 118 n = 425 1.9% 5.7% 10.5% 19.9% 13.5% 48.5% (51.5%) (100.0%) It is justified when individuals retaliate against sex offenders. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Predator Assignment) n = 79 n = 139 n = 104 n = 92 n = 38 n = 425 9.0% 15.8% 11.9% 10.4% 4.4% 48.5% (51.5%) (100.0%) 186

PAGE 187

Table 4 37 Continued Measure (Offender n = 452) Code Frequency Percent Individuals who retaliate against sex offenders should be subject to legal action. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Predator Assignment) n = 22 n = 40 n = 112 n = 172 n = 106 n = 425 2.5% 4.5% 12.7% 19.6% 12.2% 48.5% (51.5%) (100.0%) Sex offenders should be released into the community after their prison sentences. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Predator Assignment) n = 48 n = 94 n = 149 n = 125 n = 36 n = 425 5.4% 10.7% 16.9% 14.2% 4.3% 48.5% (51.5%) (100.0%) After a certain number of years, a registered sex offender should be able to be removed from the sex offender registry. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Predator Assignment) n = 95 n = 105 n = 117 n = 108 n = 27 n = 425 10.8% 11.9% 13.4% 12.3% 3.1% 48.5% (51.5%) (100.0%) It would be too harsh to make sex offenders wear a special kind of marker, at all times on their person, which identifies them as a sex offender. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Predator Assignment) n = 78 n = 78 n = 97 n = 119 n = 80 n = 425 8.9% 8.9% 11.0% 13.5% 9.2% 48.5% (51.5%) (100.0%) Having to register on the sex offender registry constitutes cruel and unusual punishment. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Predator Assignment) n = 147 n = 142 n = 79 n = 57 n = 28 n = 425 16.6% 16.1% 9.0% 6.5% 3.3% 48.5% (51.5%) (100.0%) 187

PAGE 188

Table 4 37 Continued. Measure (Offender n = 452) Code Frequency Percent I would support legislative action, which calls for sex offenders to be supervised under electronic monitoring/GPS tracking for the remainder of their lives while living in the community. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Predator Assignment) n = 41 n = 80 n = 103 n = 121 n = 107 n = 425 4.7% 9.1% 11.7% 13.8% 12.2% 48.5% (51.5%) (100.0%) I support residency restrictions that prevent all sex offenders from living too closely to schools, playgrounds and other areas where children frequently gather. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Predator Assignment) n = 14 n = 30 n = 63 n = 144 n = 201 n = 425 1.6% 3.4% 7.1% 16.4% 23.0% 48.5% (51.5%) (100.0%) Measure (Predator n = 425) Code Frequency Percent Reforms should be made to the sex offender registry. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Offender Assignment) n = 5 n = 37 n = 127 n = 176 n = 80 n = 452 0.5% 4.2% 14.5% 20.0% 9.3% 51.5% (48.5%) (100.0%) The sex offender registry is effective in reducing sex predator reoffending. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Offender Assignment) n = 27 n = 87 n = 135 n = 148 n = 28 n = 452 3.0% 9.9% 15.4% 16.9% 3.3% 51.5% (48.5%) (100.0%) 188

PAGE 189

Table 4 37 Continued Measure (Offender n = 452) Code Frequency Percent The sex offender registry makes life very difficult for sex predators living in the community. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Offender Assignment) n = 19 n = 59 n = 97 n = 180 n = 70 n = 452 2.1% 6.7% 11.1% 20.5% 8.1% 51.5% (48.5%) (100.0%) Children are safer if the locations of sex predators are known. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Offender Assignment) n = 17 n = 54 n = 77 n = 171 n = 106 n = 452 1.9% 6.2% 8.7% 19.5% 12.2% 51.5% (48.5%) (100.0%) It is justified when individuals retaliate against sex predators. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Offender Assignment) n = 82 n = 126 n = 93 n = 90 n = 34 n = 452 9.3% 14.4% 10.6% 10.3% 3.9% 51.5% (48.5%) (100.0%) Individuals who retaliate against sex predators should be subject to legal action. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Offender Assignment) n = 22 n = 51 n = 93 n = 169 n = 90 n = 452 2.5% 5.8% 10.6% 19.2% 10.4% 51.5% (48.5%) (100.0%) Sex predators should be released into the community after their prison sentences. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Offender Assignment) n = 47 n = 85 n = 129 n = 134 n = 30 n = 452 5.3% 9.7% 14.7% 15.3% 3.5% 51.5% (48.5%) (100.0%) 189

PAGE 190

Table 4 37 Continued Measure (Offender n = 452) Code Frequency Percent After a certain number of years, a registered sex predator should be able to be removed from the sex offender registry. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Offender Assignment) n = 88 n = 88 n = 109 n = 106 n = 34 n = 452 10.0% 10.0% 12.5% 12.1% 3.9% 51.5% (48.5%) (100.0%) It would be too harsh to make sex predators wear a special kind of marker, at all times on their person, which identifies them as a sex predator. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Offender Assignment) n = 53 n = 78 n = 84 n = 135 n = 75 n = 452 6.0% 8.9% 9.5% 15.4% 8.7% 51.5% (48.5%) (100.0%) Having to register on the sex offender registry constitutes cruel and unusual punishment. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Offender Assignment) n = 123 n = 130 n = 89 n = 59 n = 24 n = 452 14.0% 14.8% 10.1% 6.8% 2.8% 51.5% (48.5%) (100.0%) I would support legislative action, which calls for sex predators to be supervised under electronic monitoring/GPS tracking for the remainder of their lives while living in the community. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Offender Assignment) n = 49 n = 69 n = 101 n = 131 n = 75 n = 452 5.5% 7.9% 11.6% 14.9% 8.6% 51.5% (48.5%) (100.0%) I support residency restrictions that prevent all sex predators from living too closely to schools, playgrounds and other areas where children frequently gather. 1 = Strongly Disagree 2 = Disagree 3 = Unsure 4 = Agree 5 = Strongly Agree Missing (Offender Assignment) n = 13 n = 35 n = 69 n = 138 n = 170 n = 452 1.4% 4.0% 7.9% 15.7% 19.5% 51.5% (48.5 %) (100.0%) 190

PAGE 191

The next step in this analysis was to create two scales from the twelve offender and twelve predator measures. The two scales were created by adding the scores of the twelve measures across participants, and then by dividing that score by twelve. Once th e scales were completed then two sets of bivariate correlations were completed to see if the registry support scales were correlated with the control variables. Since there was a random assignment of "sexual offender" and "sexual predator" implemented for the registry support measures, the bivariate correlations were conducted with that random assignment in mind. In the offender bivariate correlation (Table B 41, Appendix) two variables showed significance within the correlations. Age ( r = .167, signif icant at the .001 alpha level) and education level ( r = .144, significant at the .01 alpha level) both show significant correlations with registry support (offender). The various hypotheses regarding registry support are supported at different levels. Fi rst, because two of the variables showed a significant correlation, the hypothesis regarding similarities of registry support across all social categories and groups is only partially supported (12b) Second, the parent hypothesis stating that parents w ill support the registry more than non parents is not supported due to a lack of a significant correlation (12c) Finally, the gender hypothesis which states that women will support the registry than men is also not supported due to a lack of signif icance (12d) In the predator bivariate correlation (Table B 42, Appendix), three variables show significant correlations with registry support. Gender ( r = .100, significant at the .05 alpha level), age ( r = .121, significant at the .01 alpha level) and income level ( r = .098, significant at the .05 alpha level) are all significantly correlated with registry support 191

PAGE 192

(predator). The various hypotheses regarding registry support are supported at different levels. First, because three of the variables showed a significant correlation, the hypothesis regarding similarities of registry support across all social categories and groups is only partially supported (12b) Second, the parent hypothesis stating that parents will support the registry more tha n non parents is not supported due to a lack of a significant correlation (12c) Finally, the gender hypothesis which states that women will support the registry than men is supported (12d) Next two independent sample t tests were conducted to se e if there is any significant difference between the parent and non parent groups for these scaled variables. Table 4 38 shows the results of the two t tests. The first shows that there is no significant difference in registry support between the parents (M=3.22, SD=.45) and non parents (M=3.27, SD=.39); t(450 )=1.019, p=.131 for those participants assigned to the offender language randomization. However, there was a significant difference in registry between parents (M=3.25, SD=.46) and non parents (M=3. 28, SD=.40); t(42 3 )=.535, p=.041* for those participants assigned to the predator language randomization. This provides partial support for t he parent hypothesis which predicts that parents will stronger levels of registry support, compared to their non parent counterparts (12c) The gender t tests show that there is a strongly significant difference in registry support between the male (M=3.34, SD=.43 ) and female participants (M=3.13, SD=.39); t(450)=5.198, p=.000*** for those assigned to the offender language randomization. However, there was a only a weakly significant difference in registry support between male (M=3.30, SD=.42 ) and female participants (M=3.21, SD=.44 ); t(42 3)=2.041, 192

PAGE 193

p=.042 for those participants assigned to the predator language r andomization. While both models did show a significant difference between the two genders, males scored higher in both the offender and predator randomizations. These findings do not support the gender hypothesis for this model, which suggested that wome n would support the registry more than men (12c) Table 4 38 Independent Sample T Test for Parental Status and the Registry Support (Offender and Predator Randomization). Offender Randomization Mean SD N df T Sig. Parents 3.22 .45 250 450 1.019 .131 Non Parents 3.27 .39 202 450 1.019 Offender Randomization Males Females 3.34 3.13 .43 .39 237 214 450 450 5.198 5.198 .000*** Predator Randomization Parents 3.25 .46 228 423 .535 .041* Non Parents 3.28 .40 197 423 .040 Predator Randomization Males Females 3.30 3.21 .42 .44 227 198 423 423 2.041 2.041 .042* *** p<.001 **p < .01 *p < .05 The next step in analyzing this model is to run two OLS regression to predict the participants' level of support for the sex offender registry. One model will be run on the offender randomization condition and the other for the predator randomization cond ition The additive registry knowledge variable will still serve as the independent variable and all of the previously used controls will remain control variables. However, before running the OLS Regressions the frequency statistics for a new variable m ust be analyzed. The participants perceived level of strictness has been added as a control variable for this specific model. Participants were asked about how strict they perceive their state's sex offender registry system to be. Shown in Table 4 39 t heir responses show a nearly normal distribution and center around the idea that the current laws are "Just Right." The "Just Right" response option could potentially be 193

PAGE 194

equitable to an "Unsure" option, which suggests that like the Registry Knowledge meas ures, participants are not well enough informed on accurate knowledge to be able to make a definitive response. Instead, they have chosen the neutral response option as an alternative. Table 4 39 Univariate Analysis for Perceived Level of Strictness. Mea sure (n = 877) Code Frequency Percent How strict do you think the current laws are concerning your state's sex offender registry system? 1 = Way Too Strict 2 = Strict 3 = Just Right 4 = Lax 5 = Way too Lax n = 36 n = 204 n = 353 n = 222 n = 62 4.1% 23.3% 40.3% 25.3% 7.1% Table 4 4 0 shows the OLS Regressions for both the offender and predator randomized conditions Overall both models are significant the offender regression is significant at the .001 alpha level and the predator regression is significant at the .01 alpha level indicating that the offender model is stronger than the predator model. The predictor variab les are able to account for 12.7 % (offender) and 5.0 % (predator) of the respective explained variance, as shown by the R Squar es. For the offender model, only the stereotypical offender and educational level variables are significant predictors for registry support (significant at the .001 and .05 alpha levels respectively). The parental status variable is not significant for the model this finding is consistent with the results of the first t test, which also showed no significant difference between the two participant groups for the offender randomization (12c) For the predator regression model, level of strictness (signific ant at the .01 alpha level) was the only significant predictor variable for registry support. Although it is a non significant variable, t here is a positive relationship between registry knowledge and registry, which indicates that as knowledge increases the more likely they are to 194

PAGE 195

support the registry. However, there is no significant relationship in either model between parental status re gistry support, so the parent hypothesis which states that parents will stronger levels of registry support, compare d to their non parent counterparts is not supported (12c) Gender was not a significant predictor for either model, but the negative direction of the variable remains consistent from the gender t tests. These findings do not support the gender hypothesi s as it relates to registry support (12d) Table 4 40 OLS Regression Predicting Registry Support (Random Assignment). Variable Offender Randomization Predator Randomization Coefficient Std. Error Beta Coefficient Std. Error Beta Registry Knowledge .018 .010 .086 .002 .011 .011 Stereotypical Offender .062 *** .014 .225 .021 .016 .073 Level of Strictness .035 .020 .082 .058** .023 .122 Parental Status .003 .052 .004 .017 .054 .020 Num. School Age Children .005 .026 .010 .009 .029 .017 Gender .001 .004 .007 .052 .044 .059 Age .016 .010 .092 .017 .010 .093 Race .078 .047 .088 .004 .049 .004 Ethnicity .114 .064 .083 .001 .003 .014 Education Level .082* .042 .096 .004 .048 .005 Marital Status .060 .043 .070 .067 .046 .077 Income Level .044 .027 .083 .055 .034 .091 SES Status .007 .038 .010 .037 .045 .047 Geographic Region .033 .041 .038 .056 .045 .062 Population Size .020 .016 .059 .006 .017 .016 Constant 3.777 .194 3.567 .128 F Statistic 4. 371 *** 1.476 ** R Square .127 .0 50 *** p<.001 **p < .01 *p < .05 Parental Status (dummy coded: Non Parent/Parent), Race (dummy coded: Non White/White), Ethnicity (dummy coded: Non Hispanic/Hispanic), Education Level (dummy coded: No Bachelor's Degree/Bachelor's Degree), Marital Status (dummy coded: Not Married/Not Married), Geographic Region (Not from the South/From the South) 195

PAGE 196

The registry knowledge variable did not significantly predict registry support for either th e offender or the predator model T he predator regression model weaker than the offender model and was able to explain less of the variance overall. Unlike previous regression models where the predator model was the stronger of the two, this model does n ot suggest that there is a robust difference about the terms "s exual offender" and "sexual predator" (12b). To conclude this chapter, a summary of the results has been included. Since there were numerous t tests and regressions, it was necessary to recap the findings of the ten models before moving on to the discussion chapter of the dissertation. In Table s 4 41 to 4 44 the models are abbreviated. T he overall findings and all conclusions about the various hypotheses are included. 196

PAGE 197

Table 4 41 Analysis Summary for Personal Orientations toward the Control of Sexual Offenders. Model Hypothesis and Support T Test and Significance OLS Regression and Support 1: Registry Website Access 1 a ) Generally, few people will have accessed the registry (Not Supported) 1b ) The extent to which people access the registries will be similar across social categories and groups (Not Supported by the correlations) 1c ) Parents will access the registries more often than will non parents. (Supported by t test) 1d ) Women will access the registry more often than will men. (Supported by t test). Bivariate Correlations for Registry Access and Control Variables: Number of Kids (sig .015*) Education Level (sig .032*) Marital Status (sig .013*) SES (sig .029*) Ge o. Region (sig .032*) Population Size (sig .031*) Registry Access Parent: A significant difference between parents and non parents (sig .012*) Gender: A significant difference between men and women (sig. .039*) N/A Conclusion: More community members were accessing the registry than anticipated, with a little more than half reporting that they have accessed the registry at one point. However, the proposed parent and gender differences did yield significant results, which supports the two sub h ypotheses 2: Registry Knowledge Registry Knowledge: 2a ) Level of knowledge will be low (Supported). 2 b ) Level of knowledge will be similar across social groups and categories (Supported). 2c ) Parents will have higher knowledge than non parents. (Supported) 2d ) Women will have higher knowledge than men. (Supported) Stereotypical Sex Offender: 3a ) Level of inaccurate stereotyping will be high. (Not Supported) Registry Knowledge Bivariate Correlations: No significant correlations between Registry Knowledge and the control variables T Tests: A significant difference between parents and non parents (sig .049*) A significant difference between male and female participants (sig .033*) Stereotypical Sex Offender Bivariate Correlations: Age (sig .000***) Education Level (sig .000***) SES (sig .001***) 197

PAGE 198

Table 4 41 Continued Model Hypothesis and Support T Test and Significance OLS Regression and Support 3: Stereotypical Offender 3b ) The level of inaccurate stereotyping will be similar across social categories and groups (Partially Supported). 3c ) Parents will have higher levels of inaccurate stereotyping than non parents. (Not Supported) 3d ) Women will have higher levels of inaccurate stereotyping than wil l men. (Supported) T Tests: No significant difference between parents and non parents (sig .728) for overall typology count A significant difference between male and female participants (sig .000***) for overall typology count N/A Conclusion: Overall, citizens do not have very high levels of registry knowledge. Parents and women have higher levels of knowledge compared to non parents and men, respectively. The same does not hold true for the stereotypical sex offender variable. Participants are be tt er able to identify the stereotypical offender profile compared to the registry knowledge questions, which identify more legal knowledge issues. Parental status is not significant, but gender is a significant predictor for the stereotypical offender varia ble. This difference between variables might suggest that participants are more aware of "popular knowledge" (as presented in the stereotypical offender model) compared t o "legal knowledge" (as presented in the registry knowledge model). 198

PAGE 199

Table 4 42. Analysis Summary for Fear of Victimization Model M 4 : Fear of Personal Victimization 5: Fear of Victimization of Children Fear of Personal Victimization: 4a ) The level of fear of personal victimization will be moderate. (Partially Supported) 4b ) The level of fear of personal victimization will be similar across social categories and groups. (Partially Supported) 4c ) Non parents will have more personal fear than will parents. (Not Supported) 4d ) Women will report more fear of pers onal victimization than will men. (Supported) Fear of Victimization of Children: 5a ) The level of fear of victimization of children will be high. (Not Supported by Univariate) 5b ) The level of fear of personal victimization will be similar across social categories and groups. (Partially Supported) 5c ) Parents will report more fear of victimization of children than will non parents. (Supported) 5d ) Women will report more fear of victimization than will men. (Partially Supported by t test only) Personal V ictimization Bivariate Correlation: Gender (sig .043*) Age (sig .000***) T Tests: No significant difference between parents and non parents (sig .552) A significant difference between male and female participants (sig .000***) Victimization of Children Bivariate Correlations: Parent Status (sig .000***) Number of School Age Children (sig .000***) Age (sig .008**) T Test: A significant difference between parents and non participants (sig .000***) A significant difference between male and female participants (sig .006**) Personal Victimization: R Square: 9.0%*** Significant Variables: Parental Status** Gender* Age** Race*** Victimization of Children: R Square: 10.3%*** Significant Variables: Stereotypical Offender** Parental Status*** Age*** Race** Conclusion: Registry knowledge is not significant in either model. Gender is significant for the personal victimization only this suggests that there is a gender difference in regard to victimization. Women are more afraid of becoming vict imized in than males a finding consistent with prior fear of crime literature. The parent status variable is reported in a positive direction, which indicates that parents are m ore afraid for both categories. The positive relationship fits with the predicted hypothesis regarding fear of victimization of children, but it does not fit the hypothesis regarding the fear of personal victimization. In order for the predicted hypothesis to be supported, the parent variable wou ld had to be signifi cant with a negative direction, implying that non parents are more personally afraid of victimization tha n parents. 199

PAGE 200

Table 4 43. Analysis Summary for Moral Panic Regressions Model 6 : Moral Panic Concern 6a ) Concern about sex offenses and offending will be high. (Not supported by Univariate) 6b ) Concern will be similar across social categories and groups, and it will be similar for the group responding to "sex offender" and that responding to "sex predator" survey items. (Partially Supported) 6c ) P arents will be more concerned than non parents. (Supported). 6d ) Women will be more concerned than men. (Supported). Bivariate Correlations: Offender: Parent Status (sig .000***) Number of School Age Children (sig .003**) Gender (sig .030*) Predato r: Parent Status (sig .000***) Number of School Age Children (sig .019*) Gender (sig .000***) Parent: T Test Offender: A significant difference between parents and non participants (sig .006**) T Test Predator: A significant difference between parents and non participants (sig .008**) Gender: T Test Offender: A significant difference between men and women (sig. .001***) T Test Predator A significant difference between men and women (sig. .000***) Offender Randomization: R Square: 11.9%** Significant Variables: Parental Status*** Gender* Age** Race*** Predator Randomization: R Square: 14.9%*** Significant Variables: Registry Knowledge* Parental Status*** Gender*** Age** Race** Conclusion: Both models are strongly significant, but the predator model is able to explain more variance overall and has mor e significant variables. Parental Status is significant for both models, suggesting that there is a strong level of concern among participants. This is supported by the two significant t tests related to parental status. Gender is also a significant predictor in both models, but is stronger in the predator mod el suggesting yet again that there may be a strong difference between g enders regarding these issues. There potentially could also be an interaction effect between gender and parental status in explaining the moral panic. 200

PAGE 201

Table 4 43 Continued. Model 7 : Moral Panic Hostility 7a ) The hostility/anger about sex offenses and offending will be high. (Supported) 7b ) Hostility/Anger will be similar across social categories and groups, and it will be similar for the group responding to "sex offender" and that responding to "sex predator" survey items. (Partially Supp orted) 7c ) Parents will be more hostile/angry than non parents. (Supported) 7d ) Women will be more hostile/angry than men. Supported) Bivariate Correlations: Offender: Parent Status (sig .000***) Number of School Age Children (sig .018*) Predator: Parent Status (sig .000***) Number of School Age Children (.019*) Gender (sig .000***) Parent: T Test Offender No significant difference between parents and non participants (sig .069) T Test Predator: A significant difference between parents an d non participants (sig .003**) Gender: T Test Offender: A significant difference between men and women (sig. .004**) T Test Predator A significant difference between men and women (sig. .000***) Offender Randomization: R Square: 6.7%** Significan t Variables: Parental Status** Race*** Predator Randomization: R Square: 14.5%*** Significant Variables: Registry Knowledge** Parental Status** Gender*** Race*** 8 : Moral Panic Consensus 8a ) The consensus about sex offenses and offending will be high. (Partially Supported) Bivariate Correlations: Offender: Age (sig .006**) Offender Randomization: R Square: 7.0%** Significant Variables: Registry Knowledge* 201

PAGE 202

Table 4 43. C on tinued. Model 8b ) Consensus will be similar across social categories and groups, and it will be similar for the group responding to "sex offender" and that responding to "sex predator" survey items. 8c ) Parents will be more unified than non parents. (Not Supported) 8d ) Women will be more unified than men. (Par tially Supported) Predator: Number of School Age Children (sig .028*) Gender (sig .000***) Age (sig .035*) Parent: T Test Offender: A significant difference between parents and non participants (sig .020*) T Test Predator: No significant difference between parents and non participants (sig .730) Gender: T Test Offender: No significant difference between men and women (sig. .374) T Test Predator A significant difference between men and women (sig. .024*) Stereotypical Offender** Predat or Randomization: R Square: 11.4%*** Significant Variables: Stereotypical Offender*** Number of School Age Children* Gender*** Conclusion: These models are significant overall, but do not have many individual predictor variables showing up as significant within the models themselves. Logically, this makes sense for the consensus element of the moral panic. If there was a significant difference in many of the variables, then there would not be any consensus among the participants as a whole. Parental status does not show up as a significant predictor of consensus. This may indicate that there is consensus across parental groups rather than within them. This is the third m oral panic regression model where gender is showing significance this is something that must be considered going forward as the analysis continues. 9 : Moral Panic Disproportionality 9a ) The reaction about sex offenses and offending will be disproportionate. (Supported) 9b ) The disproportionate reactions will be similar across social categories and groups, and it will be similar for the group responding to "sex offender" and that responding to "sex predator" survey items. Bivariate Correlations Offender: Parent Status (sig .001***) Age (sig .000***) Pop. Size (.014* ) Predator: Parent Status (sig .001***) Number of School Age Children (sig .016*) Gender (sig .000***) Age (sig .002**) Marital Status (sig .024*) Offender Randomization: R Square: 14.0%*** Significant Variables: Stereotypical Offender*** Age* Predator Randomization: R Square: 16.4%*** Significant Variables: Stereotypical Offender*** Gender*** 202

PAGE 203

Table 4 43. Continued. Model 9c ) Parents will react more disproportionately than non parents. (Partially Supported by t test only) 9d ) Women will react more disproportionately hostile/angry than men. (Partially Supported) Parent: T Test Offender: A significant difference between parents and non participants (sig .001***) T Test Predator: A significant difference between parents and non participants (sig .001***) Gender: T Test Offender: A significant difference between men and women (sig. .005**) T Test Predator No significant difference between men and women (sig. .068) Conclusion: For these two significant models, the predator regression is the stronger of the two yet again. Disproportionality is the most difficult of the moral panic elements to measure. The hypotheses are only partially supported because of the significant difference found between parents and non par ents within the t test, but not within the regression itself. Secondly the predator and offender t tests show a significant difference between groups at the same significance level, but the predator regression is slightly stronger than the offender regres sion. Due to these mixes results, only partial support can be given to each of the hypotheses. This is the first element of the moral panic in which gender was not a significant predictor for either model. 10 : Moral Panic Volatilit y No a priori hypotheses developed for this model Bivariate Correlations: Offender: Age (sig .048*) Ed Level (sig .037*) Predator: no significant correlations Parent: T Test Offender: No significant difference between parents and non participants (sig .147) T T est Predator: No sign. difference between parents and non participants (sig.645) Offender Randomization: R Square: 8.7%** Significant Variables: Registry Knowledge* Stereotypical Offender** Ethnicity** Predator Randomization: R Square: 5.7%** Significant Variables: Stereotypical Offender* Race* 203

PAGE 204

Table 4 43. Continued. Model Gender: T Test Offender: A significant difference between men and women (sig. .005**) T Test Predator No significant difference between men and women (sig. .146) Conclusion: These models are both significant, but neither of them show parental status as being a significant predictor at any stage. Since the models are significant, they suggest that the participants as a group do not feel as though the criminal justice system is doing enough in response to sexual offenders. Registry knowledge does show significance in the offender model, but it is only weakly significant. Th e variable is also showing a positive relationship, which indicates that as kno wledge increases so does a volatile reaction toward sex offenses and offenders. However, for this model the predator regression is not as strong as the offender regression this is the first of only two times that this change in strength has occurred. 204

PAGE 205

Table 4 44 Analysis Summary for Community Related Attitudes. Model 11 : CATSO Scale 1 1a ) Citizens will have highly negative attitudes for the overall scale and on each of the four sub dimensions (social isolation, capacity for change, severity/dangerousness, and deviancy). (supported) 11b ) These attitudes will be similar across social categories and groups. (Partially Supported) 11c ) Parents will hold more negative attitudes on the CATSO scale and sub dimensions than will nonparents. (Partially Supported) 11d ) Women will hold more negative attitudes on the CATSO scale and sub dimensions than will men. (Partial ly Supported) Social Isolation Bivariate Correlations: Age (sig .000***) Ed Level (.009**) Marital Status (sig .013*) T Test: No significant difference between parents and non participants (sig .224) A significant difference between men and women (sig. 001***) Capacity to Change Bivariate Correlations: Parent Status (sig .000***) Number of School Age Children (sig .003) Ed Level (sig .050*) Marital Status (sig .000***) T Test: A significant difference between parents and non participants (sig .000***) No significant difference between men and women (sig. .069) Severity/Dangerousness Bivariate Correlations: Parent Status (sig .018*) Age (sig .000***) SES (sig .016*) T Test: A significant difference between parents and non parents (sig .018*) A significant difference between men and women (sig. 000***) Social Isolation: R Square: 12 .2%*** Significant Variables: Stereotypical Offender*** Age** Race*** Marital Status* Capacity to Change: R Square: 12.9%*** Significant Variables: Parental Status** Gender** Race** Severity/Dangerousness R Square: 30.5%*** Significant Variables: Registry Knowledge*** Stereotypical Offender*** Parental Status* Age** Race*** Education Level* Deviancy: R Square: 12.5%*** Significant Variables: Stereotypical Offender*** Race*** Total Index of Negative Attitudes: R Square: 23.9%*** Significant Variables: Registry Knowledge* Stereotypical Offender*** Parental Status* Race*** 205

PAGE 206

Table 4 44 Continued Model Deviancy Bivariate Correlations: Parent Status (sig .002**) Number of School Age Children (sig .009**) T Test: A significant difference between parents and non participants (sig .002**) No significant difference between men and women (sig. .495) Total Index of Negative Attitudes Bivariate Correlations: Parental Status (sig .000***) Age (sig .000***) Ed Level (sig .001***) T Test: A significant difference between parents and non participants (sig .000***) A significant difference between men and women (sig. .005**) Conclusion: This model has five different components to analyze as was originally s uggested by Church et al. (2008). Overall, the results of this model show that participants largely have negative attitudes towards sex offenders. The strongest of the five models Severity/Dangerousness suggests that participant really do perceive se x offenders to be a very dangerous group of offenders, who will recidivate given the chance to. The models are overall very strong and are able to explain 26.5% of the total variance when included in the total index of negati ve attitudes. However, regist ry knowledge is not always a significant predictor. When it is significant in the model, it has a negative direction indicating that as knowledge levels decrease, negative attitudes toward sex offenders increase. 1 2 : Registry Suppor t 1 2a ) Citizens will show high support for the registry (Not Supported by Univariate)1 2 b ) The support will be similar across social categories and groups, and it will be similar for both the group responding to "sex offenders" and that responding to Bivariate Correlatio ns: Offender: Age (.000***) Ed Level (.003**) Predator: Gender (sig .042*) Age (sig .014*) Offender Randomization: R Square: 12.7%** Significant Variables: Stereotypical Offender*** Education Level* Predator Randomization: R Square: 5.0%** 206

PAGE 207

Table 4 44 Continued. Model "sex predators." (Partially Supported) 12c ) Parents will be more supportive of the registry than will nonparents. (Not supported) 12d ) Women will be more supportive of the registry than will men. (Partially supported by the t test only) Income level (.047*) Parents: T Test Offender: No significant difference between parents and non participants (sig .131) T Test Predator: A significant difference between parents and non participants (sig .041*) Gender: T Test Offe nder: A significant difference between men and women (sig. .000***) T Test Predator: A significant difference between men and women (sig. .042*) Significant Variables: Level of Strictness** Conclusion: For this model, an additional predictor variable was included the variable measured the perceived level of current strictness regarding the legal climate surrounding the registry system. It was significant for the predator regression only. Interestingly, the offender regression was stronger than the predator regression. Since parental status was not a significant predictor for either regre ssion, it is assumed that there is no large difference in support for the registry between the partici pant groups. Given everything that has been found by the previous models, this should have been a much stronger model since registry support was predicted to be high. However, it is one of t he weakest overall. 207

PAGE 208

CHAPTER 5 DISCUSSION AND CONCLUSIONS Discussion The last tables in Chapter 4 (Table 4 41 to 4 44) shows a summary of the results from all of the analyses run throughout the study. The summary shows a few key items that must be addressed. First, the use of Registry Knowledge as the primary independent variable did not yield many significant findings. When Registry Knowledge was a significant predictor, it was more frequently significant in the predator models compared to the offender models. However, it did provide the desi red negative relationship between itself and the different dependent variables. This indicates that as knowledge decreases, the respective dependent variables increase. These relationships are consistent with prior literature regarding the directionality of the knowledge variable. The secondary independent variable, the Stereotypical Sex Offender variable, yielded much more consistent, significant findings for both the offender and predator models. Second, most of the models were strongly significant overall, which suggests that the Moral Panic perspective was the correct theoretical lens to use in application to the sex offender registry. There are some issues with the individual variables within the model, but they will be discussed in later parts o f this chapter. Third, there were issues with the group comparisons between parents and non parents at the different levels of analysis. In some instances, a significant difference was predicted between the two groups for the independent sample t tests, but then the parental status variable failed to show significance within the OLS Regression. This will also be addressed later in the chapter within the discussion of a possible mediation effect with gender. Gender was a much more consistent predictor 208

PAGE 209

va riable than parental status, with women often having stronger feelings regarding sex offenses and offenders than men. Finally, there were significant finding in terms of the randomized application of "offender" and "predator" measures. This suggests that participants are able to identify the difference between the two terms, but this is such a subtle and nuanced construction that further examination of this concept needs to occur. This project started out with a theorization regarding how community member s felt about the sex offender registry and about sex offenders themselves. The results of the study largely supported the original hypothesis, which suggested that community members have highly negative attitudes regarding sex offenders and are supportive toward the use of the registry. This project also wanted to test the participants' knowledge level regarding the registry that they lend so much support to. The results showed that participants did not have very high levels of accurate knowledge pertain ing to the laws, despite expressing support for the registry. This presents an inverse relationship between knowledge and registry support (less accurate knowledge predicts increased registry support) something that reaffirms what prior literature has f ound (Proctor et al., 2002; Redlich, 2001). However, despite their inability to correctly answer questions regarding legal knowledge, participants had more success in correctly identifying an appropriate sex offender profile. That developed profile was consistent with prior literature regarding the common demographic features of registered sex offenders (Akerman et al., 2011). This brings into question the participants the type of knowledge most salient to participants "popular knowledge" vs. "legal k nowledge." This project also utilized a split in 209

PAGE 210

participation; meaning parent participants were contrasted against those participants who do not have children. This variable was not often significant, despite the hypothesis that it would be. One of the main tenants of the dissertation rested on the premise that parents and non parents would be significantly different in the way they perceive sex offenders and the registry. Despite the lack of a significant difference between the two groups, these findi ngs are important to the moral panic conversation. If there is no significant difference present, then this might indicate that participants are similar in their feelings across social groups, rather than being similar only within social groups. Finally, this study also implements random assignment of "sexual offender" and "sexual predator" questions. The purpose of this randomization was to see if participants are able to tell the difference between the terms offender and predator, or if they are equatin g the two types offender despite the fact that there is a strong difference. The predator regressions are typically stronger and more significant than the offender regressions. This indicates that participants are able to distinguish offenders from preda tors if they were not able to, then both regression models would be similar in the strength of the model's significance as well as in the explained variance. However, there was not a huge gap between the offender and predator models for any of the rando mized models. Instead, the two models only differ by a few percentage points and the total explained variance is never very large at any point in time. This concern and several other remaining issues will be discussed in the balance of this discussion ch apter, beginning with the main contributions to the field. 210

PAGE 211

Contributions to the Field Although there are several areas of this study that need to be further evaluated and analyzed, this extensive study provides multiple contributions to the field in terms of knowledge, policy, theorization, methodological conceptualization, and consensus of the topic being study. In this section of the paper, each of these contributions will be examined independently to highlight the importance of this research. Study Meth odology The first contribution present is the study's design itself. What started off as being an examination of only one state's registry, developed into a nationwide study of a very controversial topic. The inclusion of a national sample allows for a m ore comprehensive perspective regarding the sex offender registry and takes into account regional differences throughout the country. Not all states publically list offenders the same way on the registry, however participants are able to identify the nuan ces of offender classification. This suggests that across the country, it is inappropriate for them to be lumped together under the umbrella term of "sex offender." More distinctions need to be made and this study contributes to the notion that our count ry is able to make the distinction between different classification terms should a new terminology be implemented. This sample also includes a strong representation of minority participants this is not a frequent occurrence in a majority of criminologica l research. The strong minority presence allows for further examination of what we have previously known in the race and crime relationship. Race also consistently showed up as a significant predictor for many of the OLS regression models, but in an inve rse relationship. This indicates that minority participants are more cognizant of the potential threat that sex offenders pose. 211

PAGE 212

Traditionally, criminologists have recognized that most offenders offend against victims within their same racial group. This study suggests that minority participants are more afraid, concerned and angry about the possible threat that registered sex offenders pose. Most sex offenders are white and participants were aware of that fact when they were questioned on the stereotypi cal sex offender characteristics. This is an interesting finding since there is an intra racial effect occurring, rather than an inter racial one. However, despite the literature base that suggests most victimizations happen interracially (Rader, Cossman & Porter, 2012), there are still strong concerns occurring intra racially. The strong significance might be due to the higher percentage of minority participants in this study in comparison to other criminological research. But even with higher levels o f minority participants, these findings are still unexpected given what is known about race, crime, and victimization. This is a contribution to what we know about the relationship, specifically within the field of sexual offenses. There is almost an exac t even distribution between male and female participants present in this study. While it was not necessarily anticipated to be so strongly significant, gender proved to be significant in nearly all of the OLS regressions for the predator random assignment What this suggests is that perhaps female participants have differing views of sex offenders comparative to men this might be due to women having higher rates of victimization compared to men, or it might be due to the fact that women worry more about their children becoming victimized compared to fathers. Either way, a potential interaction effect may be occurring between gender and knowledge, or gender and parental status. This needs to be further addressed in future analyses. 212

PAGE 213

There is also a div ersity to the sample that makes this research relevant to all participant groups. The registry is no longer an issue that only parents pay attention to. The massive amount of media consumption associated with the registry and its laws that makes this a l arge scale issue across generations, across racial groups, and across personal boundaries. The diversity in the sample contributes to the literature on perceptions of sex offenders and perceptions of the sex offender registry, which doesn't highlight the important differences between the types of participants. There are no studies that test a parent vs non parent sample this study provides that very important test to increase the breath of information surrounding the perceptions literature. Theoretical Contributions This study also has theoretical contributions to what has been traditionally conceptualized in the moral panics literature. Cohen's work was the main focus of how moral panics were incorporated into the study. The results of the study affir med that a panic is still occurring in regards to the sex offender registry, but there were still some problems with the different elements of the panic as proposed by Cohen. Disproportionality and volatility were the weakest of the five elements and the results of those two models are consistent with Cohen's work, which suggests that it is difficult to measure these two items in real time, as the participants might not be aware of these items occurring. Disproportionality was measured using different hy pothetical sanctions that could be applied to sex offenders once they are released back into the community. Participants were then asked to address if they agreed or disagreed with this sanction being a good idea to implement. Although the explained vari ance was similar to the concern, hostility and consensus models, there were few individually significant 213

PAGE 214

variables in the disproportionality model, suggesting it to be weaker than those three. Volatility was measured by asking participants about the speed in which law enforcement, the media, and politicians react after a sex crime occurs. Participants were then asked to agree or disagree with the efficiency in reaction time for those groups. The explained variance for the volatility model was the weakest of the five elements overall, and again there was an absence of individually significant variables within the model itself. One reason for the decreased significance attributed to the volatility model may be in the rationale that volatility does not decr ease when related to sex offenders; rather volatility is a sustaining reality when it comes to this type of population. Individuals are angry about oil spills when oil spills occur, but for the most part large sections of the country are not constant angr y and upset about the possibility that an offshore oil rig could explode at any moment. The volatile reaction occurs after the rig explodes. Also oil rig disasters while sometimes catastrophic to the environment and the to the price of gas are rare i n occurrence. But sex offenses seem to be an almost daily occurrence that is broadcast on the evening news. This infers that when applied to sex offenders, volatility may never go away and instead produces something that the current state of research is ill equipped to handle. Cohen states that some of these elements such as volatility may be ongoing. Perhaps this type of element can only be captured through longitudinal research, as volatility measures the ebb and flow of public opinion. This stud y looked at a cross section reaction regarding institutional responses of law enforcement, the media and legislators once sex crimes occur. The solution to the volatility issue might also rest with the CATSO scale. 214

PAGE 215

These two elements cause concern for the overall theoretical conceptualization as proposed by Cohen. This study has had some success predicting that a moral panic still exists in regard to the sex offender registry; however there may be a better way to test for the presence of a moral panic. T he addition of the CATSO scale was fundamental to the idea that testing for moral panics can be done a bit differently from how Cohen originally conceptualized the idea. Although it needs to be flushed out more, the CATSO scale has a lot of direct links t o the measurement of moral panics as was previously discussed in this chapter. The CATSO scale models were stronger than any of the moral panic models, which suggest a better fit in terms of predicting the presence of a panic. More analysis, further test ing, and perhaps a slight reformulation of the scale is needed, but the foundation of the CATSO scale might be the exact type of scale to overcome the issues with disproportionality and volatility. This contribution to how better study moral panics could be a fundamental change to the literature surrounding the original theorization of the topic. Testing for a moral panic requires direct contribution from the community members who might be directly involved in the panic itself. Cohen states that the best way to measure a moral panic is to directly survey community members, which was done in this study. Using a survey allows for a direct measure of community member perceptions and whether or not they are able to detect that a panic is actively happening. Using regressions to examine the explained variance allows for confirmation of the panic through participants' perceptions of the event occurring. Implementation of the offender and predator measures allows for a direct comparison in terms of the panic it self for example, if the predator model is stronger than the 215

PAGE 216

offender model for the element of concern, these results suggest that participants are more concerned about the more dangerous group of sex offenders. However, it does raise the issue of vari ability in the models themselves. There are several variables that do not show significance within the regressions, such as the predicted variable of parental status. This lack of significance suggests that there is more consensus between participants wh o fall into these categories. In other words, parents and non parents are in agreement regarding specific issues tested in the study. This type of consensus does provide good results that can be contributed to the literature base itself if there is not a lot of variability among participants, then the feelings and perceptions of sex offenders are similar across groups. Essentially it is about the degree of consensus that people are exhibiting. From this dissertation it is pretty well supported that th e participants hold negative attitudes toward sex offenders. The direction of these negative attitudes is consistent across groups, but the degree to which these negative attitudes are produced might vary slightly. This in itself supports the idea that a moral panic is actively occurring regarding sex offenders and the registry, and that it is a pervasive panic that reaches across different demographic boundaries. Conceptualization This study began with the hypothesis that participant's knowledge regard ing the laws surrounding the registry would be the guiding force for a lot of their feelings and opinions. However, when inserted in the models the Registry Knowledge variable was not a consistently significant predictor variable. Instead when it was sig nificant, it was more frequently significant only for the predator models. However, the stereotypical offender variable proved to be a better predictor variable for many of the models. This 216

PAGE 217

suggests two things about the participants' knowledge base. Par ticipants do not have a high amount of legal knowledge, but they are well informed regarding the stereotypes that the media consistently portrays. This means they are knowledgeable but perhaps legal knowledge is just not the right type of knowledge for th is type of study. Testing the participants' knowledge also provides a contribution to the field because it is clear that participants are not very knowledgeable about the legal constructs surrounding a supervision tool that they are very much in support of This contribution has significant ties to the policy contribution that this study holds. For these participants it seems that popular knowledge (as shown by the stereotypical sex offender variable) is more pervasive than legal knowledge. Popular knowl edge is something that could be picked up in a moment over the evening news and is potentially more pervasive than legal knowledge. News reports often show pictures of newly arrested or newly convicted sex offenders, which provide a visual connection to t he viewer. Participants listen to these stories and their popular knowledge base increases. If their preferred knowledge outlet is the Internet, perhaps they are learning about these stories via online news or media sources. Either way, photographs are likely to accompany the story providing a secondary way for the participant to increase their knowledge base. The intricacies of law are often not important to those it does not apply to. For example, if I am a driver I am more likely to know traffic la ws compared to someone who only rides public transportation. The basic knowledge is there pertaining to things like stopping at stop signs or red lights, but all of the essentials might not be known by the bus rider. The same applies in this scenario. P articipants might be aware that sex 217

PAGE 218

offenders are forbidden from living next to places like schools or playgrounds, but because those laws do not apply to regular citizens, participants might not know the specifics regarding perimeters and boundaries. Thi s suggests that we need to find the right measures to test the right type of knowledge perhaps that is a combination of legal and popular knowledge, provided to the participants by their main source of information, the media. The right measures might al so be found by simplifying the legal knowledge measures to base form, making it more feasible that participants might have some indication that the question is correct or incorrect according to law. By making the measures too intricate, participants becom e unsure and in the end, the measures are not as useful to testing these issues. This reformulation of the registry knowledge variable might also help to increase the significant of the variable as a significant predictor in the models. Discussion of Spec ific Issues Registry Knowledge as an Independent Variable Registry Knowledge was used as the primary independent variable in all of the OLS Regression models. It was not always a significant variable in the models, and it was not always a significant pred ictor for both the offender and predator randomizations. More commonly, registry knowledge was only a significant predictor for the predator models. This might suggest that knowledge has seeped into our culture and we are more attuned to these types of i ssues compared to generations past. The vast media exposure centered round potential victimizations, and the feared stranger sex offender has made us more cognizant of the fact that sex offenders should be perceived as a constant threat. Registry Knowled ge is a weakly significant predictor variable for several of the moral panic elements concern (predator model), hostility 218

PAGE 219

(predator model), consensus (offender model), and volatility (offender model). In contrast, the Registry Knowledge variable held up strongly in the CATSO Scale regression model for the construct of Severity/Dangerousness, and was significant in the Total Index of Negative Attitudes toward sex offenders as well. As related to the actual moral panic models, Registry Knowledge was not as successful a predictor as was originally hypothesized. For example, Registry Knowledge is only a weakly significant predictor variable for the Concern and Hostility, but only for the predator ones. In contrast, the Stereotypical Sex Offender variable is significant in the Concern (Predator model), Consensus (Offender and Predator models), Disproportionality (Predator and Offender models), and Volatility (Offender model). Finally, Registry Knowledge did not even provide a significant relationship with re gistry support. These mixed results create issues within the dissertation and make future research a necessity to elaborate on how to 1) correct the Registry Knowledge variable, 2) combat the problems with the moral panic measures and 3) to rectify the di sjunct between the moral panic measures and the CATSO scale. Although the Registry Knowledge variable was not consistently significant, the inclusion of the Stereotypical Sex Offender variable provided additional insight into the type of knowledge that p articipants knew. The variable was set up so that participants would be tested on what the most common characteristics came to mind regarding sex offenders. Would the myth that most sex offenders are older, white men hold up? Or would the participants b e able to create an offender profile that was closer to what prior literature has stated. By being able to correctly identify whom the most frequently registered sex offender really was, the stereotype falls away and a correct offender 219

PAGE 220

profile takes its p lace. Although participants are able to bypass stereotyping the sex offenders in terms of demographics, the title for this variable still remains "Stereotypical Sex Offender" because of the way it tested the participants. The Stereotypical Sex Offender v ariable was much more consistent than the Registry Knowledge variable, which suggests that perhaps legal knowledge is not the right type of knowledge to be measuring. The popularized media highlights sex offenses regularly and this type of information mig ht be more likely to stick in the minds of the participants. Registry Knowledge variable is often providing the hypothesized inverse relationship with the various dependent variables, but additional analysis needs to be conducted in order to find a more co nsistent and significant independent variable. One of the ways in which this could occur would be to manipulate the coding structure of the response options. Participants were given a five point Likert scale of responses to choose from, including a neutr al "Unsure" option. The "Unsure" option was utilized but a majority of the participants for several of the ten registry knowledge measures. In an attempt to overcome that hurdle, the responses were dichotomized into "Correct" and "Incorrect" response opt ions before they were transformed into a 10 measure count variable. However, this transformation did not seem fruitful in terms of producing a significant predictor variable. Perhaps a more structured scale variable like the one used in the pilot study w ould be a better fit. This was the intended format for the Registry Knowledge variable to begin with, but the factor analysis produced three non conforming measures, which in turned produced low internal validity as exhibited by the Cronbach's alpha. Fut ure analysis will try to combat this issue with the factor analysis 220

PAGE 221

to see if a different combination of items would eliminate the overall issues with the Registry Knowledge variable. The Role of the Media as Related to Moral Panics One of the main proposi tions listed in Cohen's original work on moral panics focuses on the role of the media and how their influence makes moral panics more salient to citizens because of the increase of information being disseminated. As this study's instrument was developed, it was hypothesized that registry knowledge would be very influenced by media reports. Questions were even posed to the participants regarding their information sources. The most frequent information source was the Internet which is often dominated by many different news outlets. In the era of 24 hour news reporting, it is often difficult to avoid the media and instead many people find themselves oversaturated by news coverage. That oversaturation has the potential to lead to misinformation being deli vered to the viewers, and specific biases could sway the viewers in one direction or the other depending on what channel the news is coming from. The media is running a business and news stations are better able to sell stories about child sexual abuse co mpared to adult victimizations. This causes them to highlight the stories with child victims more frequently, and in turn adds to the fear that parents have regarding the chances that their children might be more likely to be victimized compared to them. Furthermore if I am childless, but don't put myself in risky situations where becoming a victim of a sex offense is a possibility, I might not even be afraid for myself because I can protect myself. This lack of personal fear increases the amount of fear that I might have for children becoming victims of sexual crimes. In order to examine fear more closely, it was thought that prior victimization would have been a predictive factor. However, there 221

PAGE 222

was not enough reported prior victimization found among participants to hold the relationship between prior victimization and fear of sexual victimization. That relationship was dropped and the variable was removed from the multivariate models. Because this was such a large part of the build up of the study it was necessary to put the media information source in the models as a control variable. However, when this variable was put into the model it dragged the regression down. Therefore it was necessary to remove it to maintain the strength of the regression. The most frequently identified source of information was the Internet compared to print sources or TV. There might be a connection between the Registry Knowledge, Stereotypical Sex Offender, and media source variables; this possible relationship might l end support for the notion of legal knowledge vs popular knowledge, which will be discussed later in this chapter. Younger Participants Mean a Different Knowledge Base The participant sample identified as belonging to a younger age group in total, most p articipants identified as being between the ages of 18 39. Having this type of demographic would suggest that younger participants might be getting their information from sources that are different from older participants. Younger participants are more t echnology driven and it would make sense that their information sources are following suit. Many of the regression models showed age as a significant control variable, but in an inverse relationship. This means that younger participants are more likely t o exhibit more negative attitudes toward sex offenders, have stronger levels of support for the registry, and are more in tune with the moral panic issues. Like the registry knowledge variable, a generational shift might be taking place in regard to medi a attention fueling the moral panic. Younger participants stay glued to 222

PAGE 223

the Internet, which is able to provide them information at a moment's notice. Older participants generally do not have that same connection to technology that younger participants do The majority of participants are younger than 39 years old, which also contributes to the significance of the variable. Moral Panics as a Theoretical Lens Cohen's proposed Moral Panic was a successful fit to explain perceptions of sex offender and of th e sex offender registry. The models derived from the five elements concern, hostility, consensus, disproportionality, and volatility were all strongly significant and suggested that a moral panic is still taking place in regard to those issues. The r andomization of the offender and predator measures was utilized in testing the elements of the moral panic. All of the predator models were stronger than the offender models in this section, with one exception lying in the volatility model. In that model the offender model was able to explain 0.4% more variance than the predator model. These results show that the moral panic paradigm was appropriate to apply to the sex offender registry. However, there is still the issue of two elements of the moral pan ic disproportionality and volatility that need to be addressed. Cohen originally stated that disproportionality would be the hardest to measure by nature of participant self report surveys that tried to measure the panic in real time (1972; 2004). Th is difficulty was also present in this study. The disproportionality regression models were only partially supported and only had minimal variables significant within the model. Although both the offender and predator models were significant overall, mor e analysis needs to be done to find the best fit for prediction purposes. The volatility models were the weakest models overall for the moral panic elements. This was the first time that 223

PAGE 224

the predator model was weaker than the offender model in terms of s ignificance and explained variance. Again, this concept is difficult to measure through self report surveys. For these measures, hypothetical punishments and sanctions were suggested for participants to gauge in terms of severity. There is also a concer n that volatility might be addressing something beyond what these measures were able to capture for this study. Cohen's theorization of moral panics suggests that both disproportionality and volatility tend to shine when the panic is reinvigorated through when another offense occurs (1972; 2004). The survey used for this study was not addressing any specific offense, but rather asked participants about the panic in more of an abstract sense. It might be more useful to try to measure these two constructs i mmediately after the panic is reinvigorated and the survey questions can be geared toward a specific crime. Finally, the CATSO Scale has to be addressed in the context of the moral panic theme present in this dissertation. The CATSO scale fits well within the overall scope of the study because at its base level it measures participants' attitudes toward sex offenders, whether they are positive or negative in nature. However, breaking down the measures that are present it is clear that many of them fit wit hin the elements of the moral panic. For example, specific measures such as "People who commit sex offenses should lose their civil rights," or "Sex offenders should wear tracking devices so their location can be pinpointed at any time," both fit in the C apacity to Change component of the CATSO scale. Going further, those same two measures could just as easily serve as the hypothetical punishments used to measure disproportionality within the moral panic questions. Many of the CATSO scale measures can fi nd homes within 224

PAGE 225

the moral panic elements, except for those that comprise the Social Isolation construct. Since the CATSO measures were a stronger predictor for many of the issues being discussed in the moral panic models, it begs the question, which is th e better test? The moral panic measures produced significant models, but they simply weren't as strong as the ones in the CATSO scales. Since there are issues with the Social Isolation construct, perhaps melding the CATSO and moral panics together would produce a better way to test these issues. Many of the measures have a direct connection with one another and additional testing could find the right combination to provide a strong scale that overcomes the weaknesses of the moral panic elements and also provides a solution for the Social Isolation issue. The regression model for Social Isolation construct was a somewhat identifiable outlier in comparison to the other three constructs. The Social Isolation regression was able to explain the least amoun t of variance, and was the only model to be unable to show a significant difference between parents and non parents in the t test. This suggests that the Social Isolation model operates differently and is not able to show a direct correlation into testing the moral panic. Of the other three models Capacity to Change, Severity/Dangerousness, and Deviance parental status is significant for both Capacity to Change and Severity/Dangerousness. Despite being a significant predictor, the parental status var iable suggests that consensus is not present for parents and non parents regarding those two issues. However, because the parental status variable is not significant for Social Isolation or for Deviance this suggests that there is consensus between parent s and non parents for this issue. The moral panic model testing 225

PAGE 226

consensus also produces a non significant parental status variable, suggesting that parents and non parents are in consensus overall. This mixture of results suggests that there are some is sues in which the two groups are in agreement, but then they disagree on some other issues regarding sex offenders. It is interesting the Capacity to Change and Severity/Dangerousness are the two that do show a significant difference between parents and n on parents. This could be feeding into fear of victimization whether that is for themselves or for their children or it could be feeding into the knowledge that surrounds these issues. Regardless, it is clear that consensus is not consistently presen t between parents and non parents. However, race seems to be a more significant variable for consensus. Race was a significant predictor in all five of the CATSO models, providing a negative relationship in all of the models (minority participants have more strongly negative attitudes toward sex offenders). Although there is not consensus between minority and white participants regarding these issues, it is a significant finding that this variable was strongly significant regarding the viewpoints of min ority participants. These findings might not show consensus across racial groups, but they do show consistent consensus within minority populations. Since most sex offenders are white, these results contradict the typical research findings regarding inte r and intra racial victimization concerns. This provides a strong starting ground in terms of analysis, but further analysis needs to be done to see why the Social Isolation construct seemingly stands apart from the other constructs within the entire CATSO scale. Offender vs. Predator Distinctio n The inclusion of the randomized offender and predator measures was a unique addition to the study, which has not been seen in prior literature before now. This 226

PAGE 227

contribution to the field will expand the literature surrounding perceptions of sex offenders The two terms are somewhat similar in nature and arguably often interchangeable among the general public. Sex predators are lumped under the umbrella term of sex offender, but it is important to make sure that the two do not remain equated with one ano ther. In other words, all sexual predators are sexual offenders but all sexual offenders are not sexual predators. Even though there are subtleties between the two terms, this study shows that participants are able to distinguish the two from one another The randomized use of the offender and predator measures was implemented in the elements of moral panic, and for the registry support questions. They were not implemented anywhere else because the other measures asked primarily about the registry, or w ere replicated measures (such as the CATSO scale) and the measures needed to remain true to the original development. When it was not implemented, all of the measures used the term "sex offender" within the question. Participants were exposed to only on e of the two measures within their assigned instrument. Individual analysis was conducted for the offender and predator measures for each of the predicted models. Those individual analyses were able to show whether or not the participants were able to te ll the difference between the two terms. For the most part, the predator measures produced stronger models in terms of explained variance, overall significance within the model, and more individually significant variables. If the participants were not ab le to distinguish one term from the other, the models would be more closely similar. The issue with the offender and predator research is the inconsistencies with the registry knowledge variable. There is a legal 227

PAGE 228

difference between the two terms and the predator is certainly more severe in terms of the sex offenses committed. Yet again, the stereotypical sex offender variable is a better predictor for these randomly assigned measures within the models for the five elements of moral panic. This might be an issue of the type of knowledge that is gathered by the participants, and an explanation might simply rest in the words used. Predator implicitly has a more negative connotation attached to it, suggesting that there is more reason for the moral panic to exist with this group of individuals compared to the offenders. The manipulation of the current registry knowledge variable, or perhaps a brand new independent variable will help to explain the nuances between these two terms on a better level. Limitat ions Although this project has many merits and strengths, there are some limitations that must be acknowledged. First, the use of the online participant pool MTurk has several disadvantages. MTurk is a voluntary, nation wide participant pool, but because it is voluntary there may be some selection bias occurring. Furthermore, this bias might be increased because there was a $1 monetary incentive. This is a rather a large incentive compared to the other surveys that were being administered around the sam e time. Most of the surveys were only providing $.25 or $.50 to participants as an incentive. Since the participants tended to be younger in age, this might imply that there might be a generational bias in participation as well. Younger participants are more reliant on the Internet and might be using MTurk as a way to supplement their income. The $1 incentive would certainly be more appealing to a younger demographic comparative to an older demographic group. 228

PAGE 229

Although this project was originally intende d to only survey residents in Florida concerning the Florida Sex Offender Registry, the use of MTurk immediately turned a one state study into a nationwide study. Although at least one participant from every state did take the survey, making it national i n scope. However, despite this fact the sample itself is not representative of the country as a whole. Several states stood out with higher rates of participants and a concentration of southern states was more strongly represented than other parts of the country. The South has traditionally had more punitive laws and more punitive views toward the punishment of offenders, which may have biased the results of the survey somewhat. Since the sample is not representative of the country overall, this has imp lications for future research. Specifically, this sample could not be used to test a moral panic exactly after a new federal sex offender law was passed. However, it might be possible to test regional differences if more participants could be pulled from specific areas of the country. Since the survey quickly developed into a nationwide study, the survey instrument had to be revised to test participants regarding their knowledge of federal laws regarding the sex offender registry, rather than the state la ws that were originally going to be used when the survey was still being applied only to Florida. This is problematic since many states have different registry systems and are not required to have identical structures. Federal law requires that states ma intain the minimum legal standards in order to be in compliance with the government, but beyond that states have a lot of legal leeway in how they chose to supervise their resident sex offenders. Participants might be better acquainted with their own stat e laws rather than the federal statutes that were used to test the accuracy of their knowledge. However, this was the 229

PAGE 230

only viable option to use since the survey was administered to be participants all over the country. Testing participants on their indiv idual state laws might increase the accuracy of their knowledge, but this would involve either the administration of fifty one (to account for the District of Columbia) different randomized surveys or fifty one individual projects. Finally, there is a bit of unevenness when it came to the administration of the offender and predator randomizations. Qualtrics automatically accounts for randomization by evenly assigning participants to each group. So in theory, the participant groups should be absolutely eve n. However, one of the functions of MTurk allows for the researcher to supervise the responses of the participants to make sure they are not just answering the first response they find and that the responses were somewhat reliable at face value. Based on the amount of time it took participants to take the survey, those who took less than 5 minutes were reviewed to make sure the responses were reliable. Several responses needed to be thrown out, thus making the perfect random assignment uneven. Although the ideal situation would keep that perfect evenness in tact, there is not such a large difference (only about 25 participants) in offender and predator responses that this is a serious concern. Additionally, there may be a contamination effect occurring in regard to the "sexual offender" and "sexual predator" randomization. The randomized questions appeared toward the beginning of the survey, were applied in the middle of the survey, and also at the end of it. By receiving those randomized measures, pa rticipants might have been swayed by the use of the terms offender or predator. This contamination effect could potentially account for the stronger predator models, and also for the lower 230

PAGE 231

sample of participants who were assigned the predator measures. T here may be something in the predator language that turned participants off and accounted for fast responses or attrition. Future Research Although Cohen's moral panic was used as a theoretical lens for this dissertation, there are other theoretical explan ations that could be applied to research on perceptions of sex offenders and the sex offender registry. Labeling Theory and/or Reintegrative Shaming Theory both provide sufficient theoretical bases to apply to a legal structure such as the sex offender re gistry. There is a lot of variability in terms of sex crimes that could require a convicted sex offender to register with the state. In addition to that variability, there is also a gray area as to whether or not all of these offenses should be sentenced the same way. Activities such as public urination, engaging in sexual intercourse, skinny dipping, mooning or streaking have the potential to land someone on the sex offender registry due to the exposure that comes from these actions. These are some act ivities where one could argue there is a little bit of a gray area as to whether or not these individuals are really sex offenders. By definition they could be, but they are not viewed in the same way as many other sexual offenders and predators. The poi nt is that while legally these individuals might be treated the same in the eyes of the law, there is a dualistic fallacy taking place (Reid, 1982) in the minds of the normative citizens there is no gray area because sexual offenders are an inherently di fferent group of people, regardless of the fact that legally they could be labeled the same. Secondly, Braithwaite's theory of Reintegrative Shaming describes the idea that sanctioning someone for their crimes brings about shame for the individual, but aft er the shaming occurs there needs to be efforts made to bring the shamed individual back into 231

PAGE 232

the fold (1989). The shaming can be disintegrative when there is no effort made to rehabilitate the offender and reincorporate the individual back into society. R esearch has shown that the registry is more disintegrative in nature and keeps sexual offenders on the fringes of society (Klein, Zambrana and Rukus, 2012; Tewksbury, 2004; Tewksbury and Lees, 2007). Reintegrative Shaming Theory would suggest that due to t hese hardships and the difficulties associated with reentry, the potential for recidivism would increase (Braithwaite, 1989). Ironically, this is the exact opposite goal of the registry and the sex offenders that have been passed to keep sex offenders in c heck (Klein et al., 2012). This brings up an important point within the perceptions literature. Even though community members are highly supportive of the registry and the supervision of sexual offenders, how many citizens legitimately know about the conse quences of the registry? Or are they so invested in the idea that all sexual offenders are getting their just desserts that they are uncaring about difficulties with re entry? Issues such as housing, employment and vigilantism are all concerns for sex offe nders returning to their communities. For those citizens who are supportive of the harsh registry laws, it is suggested that these efforts are backfiring and potentially are setting the stage for an increased level of sexual offender recidivism. Reintegra tive Shaming provides a unique perspective to the post release issues that often accompany the reentry efforts of registered sex offenders. There are multiple avenues of research in this study that could be extended in future studies. First, researching S kolnick's "Symbolic Assailant" in regard to identifying the stereotypical sex offender was included in this survey. However, this is something 232

PAGE 233

that needs to be taken further. A vignette study has been developed to examine the offender profile of individu als engaged in sexually deviant scenarios to determine which person fits the stereotypical mold of a sex offender. This type of study could manipulate the race of the sex offender to see if the significant race findings of this dissertation could be repli cated elsewhere. This could also be an avenue for testing the significant gender findings as well. Manipulation of the offender's gender might also provide a way to better understand the consistent gender findings that were also present in this dissertat ion. Finally, in order to see if the moral panic is truly pervasive on a national level, this survey should be administered a second time to see if there is a large change in participant attitudes and concerns following a high profile sex offense. For ex ample as discussed in the literature review, many of the federal laws that were passed regarding the sex offender registry were approved by legislators after a high profile offense like Jacob Wetterling or Megan Kanka took place. It would be an interestin g study to see how people felt about sex offenders before and after those events took place. This survey data could certainly serve as the before portion of that type of study and the re administration of the survey could serve as the after portion. A co mparison of the data could then occur to see how much influence an event like that would have on participant attitudes, beliefs and knowledge regarding sex offenders. While this is not a desire for a horrible crime to take place, it is certainly something that could be administered should an even happen. Policy Implications As stated earlier in the paper, moral panics are not a new phenomenon by any means. They have the ability to sweep the country, or can be very small in nature. 233

PAGE 234

They can also be fleeti ng or long lasting occurrences. Cohen states that, "Sometimes the panic passes over and is forgotten, except in folklore and collective memory: at other times it has more serious and long lasting repercussions and might produce such changes in legal and social policy or even in the way society conceives itself" (2004:1). The long last repercussions, and legal and social policy chances are the most concerning for the atmosphere surrounding sex offenders. There are a variety of policy implications that c an be derived from this study. The first, and maybe most important, is that the nation does not know enough about the laws that they are highly supportive of. Through the Stereotypical Sex Offender variable, the study shows that participants are more awa re of who sex offenders look like and the types of offenses they are committing. But their legal knowledge is still limited. The inconsistencies with the Registry Knowledge variable are problematic to the study, but this type of knowledge is arguable the more important type because the law is what changes not the type of offender. If legal knowledge is already limited, it is possible it will decrease in accuracy as the laws surrounding the registry evolve and expand. Despite the inconsistencies within the model, simple frequency statistics reveal that the many participants either were unsure or completely wrong about the information presented to them. However, their support for the registry, and even harsher penalties against sex offenders was as stro ng as prior literature suggested it would be. For the most part, the laws surrounding the sex offender registry are emotionally charged, but not very effective in preventing sexual offenses. We get our information from the evening news, which is trying to sell stories, so they heighten the language surrounding their reports and making it emotionally driven. They discuss that law 234

PAGE 235

enforcement has arrested a "sex predator" and then go on to describe this individual as someone with no prior arrests or convi ctions. This doesn't make that person a sex predator, it makes him or her a person who was arrested on allegations of a sex crime. Once that term is broadcasted, that is all that anyone will associate with that individual. However, the country still love s to support them because they are geared toward curbing the behavior of a generally abhorred group of offenders. The moral panic surrounding sex offenders often leads to increases in punitive sanctions, translating to longer prison and registry sentence s. Just as we saw the panic have a hand in the formation of the modern day registry system in the late 1980s and early 1990s, we may be about to see that happen again. Although we haven't seen any recent overarching legal change at the federal level, the re are some states that are beginning to become more punitive towards sex offenders and their states' registry systems. In Florida, there is a strong push toward harsher penalties and more transparency with registry information. Most recently, Florida h as included a more specific search mechanism on the registry website provided by the FL Department of Law Enforcement. Prior to this new mechanism, individuals could search registered sex offenders could based on neighborhood (address) searches, email add ress/Instant Messenger IDs, or by the offender's name. These searches would generate a list of registered sex offenders who met the search criteria. The newest addition to the website allows for individuals to search for sex offenders on university campu ses whether they are students, employees, or volunteers of the school. This includes anyone who is registered in Florida and some affiliation with the schools but interestingly this extends beyond 235

PAGE 236

schools in Florida, and also includes online programs. The point of the search is to provide information on sex offenders who might be a potential danger to those community members who are physically on the same campus as sex offenders. There are a variety of online schools, out of state schools, and even a pastry school included in eligible universities. While this certainly makes the criteria comprehensive, logic dictates that on campus students are probably not at risk from online students. Nonetheless, this is just one additional measure implemented by the state of Florida to keep track of its sex offenders. It was announced in September 2013, that the Florida Senate and House of Representatives would revisit the state's laws regarding punishments for sex offenders after eight year old Cherish Perrywin kle was found dead in June (Farrington, 2013). It was suspected that 57 year old Donald Smith, a registered sex offender, was at fault for her death. Smith was on the registry for a felony conviction related to the 1993 kidnapping of a young girl. After serving out an incarceration sentence unrelated to the kidnapping, he was released into the community and roughly one month later befriended a young, single mother and her children. Smith promised to take the family to a local Walmart to help the mother, Rayne Perrywinkle, buy her children some essential items such as clothing and food. Although she felt strange about the situation, Rayne Perrywinkle, accepted Smith's offer and the group went to Walmart. Once they got to the store, Smith isolated Perryw inkle's youngest daughter Cherish and allegedly kidnapped her from the store. Cherish's body was found the next day in some woods outside of Jacksonville and Smith was charged with her kidnapping, murder, and sexual battery. He has since plead not guilty and is currently awaiting trial (Pantazi & Treen, 236

PAGE 237

2013). Almost immediately after Perrywinkle's body was found and Smith was arrested and charged for her death, proponents of the sex offender registry began to call for sweeping reforms to the registry s ystem which would include increased sanctions against sex offenders. If the legal climate continues to become more stringent and punitive against sex offenders, then the unintended consequences will only continue to increase as well. Since this is only a cross section survey, it is impossible to make the conclusion that the laws will become more punitive in nature. However, the participants in this study are vocal in their support for a harsher and more stigmatizing registry system. There is a well fou nded literature base which suggests that although they were not written into law, there are a variety of side effects to current registry laws that cause sex offenders difficulties in successful re acclimation to society. Although this dissertation does n ot address unintended consequences specifically, the instrument does address whether or not participants would support increased sanctions against sex offenders. Participants did support the implementation of items such as mandatory GPS monitoring, increa sed prison sentences, and longer registration terms. If the sanctions against sex offenders continue to increase as it has in the past, then prior literature would lend itself to hypothesize that these unintended consequences would also increase. Due to residency restrictions, sex offenders often cannot live near family members, in safe areas of cities, sometimes have restricted access to public transportation (Levenson and Cotter, 2005; Levenson et al., 2007; Minnesota Department of Corrections, 2003; Te wksbury and Zgoba, 2010; Zanbergen & Hart, 2006). Having to identify as a registered sex offender has also been related to issues with unemployment (Jenkins, 237

PAGE 238

2004; Klein et al., 2012; Levenson et al., 2007; Tewksbury, 2004; 2005; Zevitz & Farkas, 2000), a nd stigmatization (Burchfield & Mingus, 2008; Zevitz & Farkas, 2000). Instead of receiving assistance to return to their communities and successfully reenter their former lives, sex offenders are frequently stigmatized and forced to live on the outskirts of communities, all in order to remain compliant with sex offender registry laws (Levenson & Cotter, 2005; Tewksbury, 2005; Tewksbury & Lees, 2006). While it is understandable that many individuals do not want to live in close proximity to sex offenders b ecause they are afraid or angry about their crimes, the social reality that we are facing is that this constant exclusion pushes sex offenders to the boundaries of society and is causing problems where they need not occur. Unintended consequences are an a rea of research that needs to be continually explored as the laws continue to increase in severity. 238

PAGE 239

APPENDIX A PILOT STUDY INSTRUME NT Research Instrument (Quantitative) The following survey will ask you questions about your perceptions of sex offenders who may be living in communities near you. You will be asked questions about how much sex offenders concern you in terms of safety, how you feel about the sex offender reg istry and your reaction to sex offender registry laws and your knowledge about the sex offender registry. As stated in the informed consent, you may stop taking the survey at any time and you may skip any question that you are not willing to answer. (T his section of the survey will include demographic questions regarding the participant). 1. What is your gender? Male Female 2. How old are you? 18 20 21 23 24 26 27 or older 3. What is your race? Black White Other 4. Do you identify yourself as Hispanic? No Yes 5. What is your class standing? Freshman Sophomore Junior Senior 6. Please indicate if you have been a victim of any of the following crimes. Property Crime (such as theft or breaking & entering) Drug Crim e Personal Crime: Non Sexual (such as robbery, assault or battery) Personal Crime: Sexual (such as rape, sexual assault or molestation) Other (This section of the survey will address participant knowledge of sex offenders that may be living in nearby neighborhoods, as well as participant knowledge of the registry rules). This portion of the survey asks you about your knowledge of the sex offender registry rules and of sex offenders that may be living in your neighborhood. 7. How many times have you looked at the Florida Department of Law Enforcement, Florida Sex Offender Registry website? 0 times 1 time 2 times 3 times 4 or more times 8. How many times have you looked at the website to search the areas surrounding your home for sex offende rs living nearby? 0 times 1 time 2 times 3 times 4 or more times 9. How many sex offenders are you personally acquainted with? 239

PAGE 240

0 1 2 3 4 or more 10. Do you know that there are two types of registered sex offenders in Florida sexual offenders and sexual predators? No Yes This isn't true 11. Do you know that registered sex offenders are required to live at least 1,000 feet from a school zone, park or bus stop? No Yes This isn't true 12. Do you know that some sex offenders are require d to register for life? No Yes This isn't true 13. Do you know that not all sex offenders are required to be on some sort of electronic monitoring/GPS tracking device? No Yes This isn't true 14. Where did you learn the most information regarding sex offenders and the sex offender registry? Television Newspaper Magazines Radio Internet School Friends/Family Other (Please Indicate): ______________________ Read the following statements and answer with the most appropriate answ er. 15. Everyone who has ever been convicted of a sexual offense is required to register on the Florida Sex Offender Registry. Very True Somewhat True Unsure Somewhat False Very False 16. The definition of a sexual predator is: Repeat sexual offenders, sexual offenders who use physical violence, and sexual offenders who prey on children are sexual predators who present an extreme threat to the public safety. Sexual offenders are extremely likely to use physical violence and to repeat their off enses, and most sexual offenders commit many offenses, have many more victims than are ever reported, and are prosecuted for only a fraction of their crimes. Very True Somewhat True Unsure Somewhat False Very False 17. Individuals convicted of their very first sexual crime can be classified as a sexual predator on the Florida Sex Offender Registry. Very True Somewhat True Unsure Somewhat False Very False 18. Juvenile offenders can be required to register on the Florida Sex Offender Registry if con victed of a sexual offense. Very True Somewhat True Unsure Somewhat False Very False 19. After serving their prison sentences, sex offenders can be incarcerated indefinitely through the process called "Civil Commitment." Very True Somewhat True Unsur e Somewhat False Very False 240

PAGE 241

20. In Florida, there are more male sex offenders registered than female sex offenders. Very True Somewhat True Unsure Somewhat False Very False 21. Female sex offenders make up roughly one fourth of the total amount of registered sex offenders. Very True Somewhat True Unsure Somewhat False Very False 22. The Amber Alert System is named after the color amber, at no time was the system designed in memory of a child. Very True Somewhat True Unsure Somewhat False Very False 23. Sex offenders have one of the highest recidivism or re offending rates of any offenders. Very True Somewhat True Unsure Somewhat False Very False (This section of the survey will start asking questions that have been formed from the mo ral panics literature adapted to sex offenders and the sex offender registry). This section of the survey asks you about your personal worries, concerns and feelings regarding the registry and the sex offenders that may be living near you. Please select the most appropriate answer that describes your point of view. (Concern) 24. Are you worried about sex offenders living nearby your home? Not really A little Somewhat A lot 25. Are you worried that you personally may be a victim of a sexual offense ? Not really A little Somewhat A lot 26. Are you concerned about sex offenders being on the college campus? Not really A little Somewhat A lot 27. Are you worried that if sex offenders are living in the community, then more sexual offenses will occur? Not really A little Somewhat A lot (Hostility) 28. Are you angry that sex offenders are allowed to live in the community? Not really A little Somewhat A lot 29. Do you feel any resentment over the fact that some of your neighbors may be se x offenders? Not really A little Somewhat A lot 30. Do you feel any anger towards the criminal justice system for releasing sex offenders from jails and prisons? Not really A little Somewhat A lot 241

PAGE 242

31. Are you angry that sex offenders may be working at businesses where you may frequently shop or visit? Not really A little Somewhat A lot (Consensus) 32. Do you feel that a majority of community members are in agreement about the risk that sex offenders pose? Not really A little Somewha t A lot 33. Do you feel that many community members feel that changes must be made in the supervision of sex offenders? Not really A little Somewhat A lot 34. Do you feel that community members in general feel threatened by sex offenders as a group? Not really A little Somewhat A lot 35. Do you feel that a majority of community members are in agreement that children are at risk of being sexually victimized? Not really A little Somewhat A lot 36. Do you feel that many parents feel that sex offenders are too dangerous to be living in the community? Not really A little Somewhat A lot (Disproportionality) 37. Do you feel that the current state of the sex offender registry is too harsh? Not really A little Somewhat A lot 38. Do you feel that the sex offender registry laws should be stricter? Not really A little Somewhat A lot 39. Do you feel that keeping sex offenders on electronic monitoring/GPS tracking for more than 5 years without a break is too severe a punishment. Not real ly A little Somewhat A lot 40. Do you feel that sex offenders should report to law enforcement more than the required two times per year? Not really A little Somewhat A lot 41. Do you feel that the media overreacts in their reporting of sex offen ses when they occur in a community? Not really A little Somewhat A lot (Volatility) 42. Do you feel that law enforcement reacts quickly enough when a sexual offense takes place? Not really A little Somewhat A lot 242

PAGE 243

43. Do you feel that legislators work fast enough to get necessary registry laws passed to further keep track of sex offenders? Not really A little Somewhat A lot 44. Do you feel that the media reports sex offenses cases too quickly before all of the facts are gathered? Not really A little Somewhat A lot 45. Do you feel that the quick response of the media makes communities safer because people are made aware of the sex offense? Not really A little Somewhat A lot 46. Do you feel that police are too slow to catch sex offende rs when sexual offenses take place? Not really A little Somewhat A lot (The next part of the survey will ask questions about participant reactions to sex offenses in general and about the rules associated with the registry.) This portion of the survey asks you about your reaction to sexual offenses in general and about the laws associated with the registry. Please select the most appropriate answer that describes your point of view on these issues. 47. Do you support the use of the publicly av ailable Florida Sex Offender Registry? Not really A little Somewhat A lot 48. How often have you heard of sex offenders being hurt or killed by community members as a result of their sex offender status? 0 times 1 time 2 times 3 times 4 times o r more 49. How often have you seen fliers, notes or posters advertising the presence of a sexual offender who is living within your neighborhood? 0 times 1 time 2 times 3 times 4 times or more 50. Reforms should be made to the Florida Sex Offender R egistry. Strongly Disagree Disagree Somewhat Unsure Agree Somewhat Strongly Agree 51. The registry is effective in reducing sex offender re offending. Strongly Disagree Disagree Somewhat Unsure Agree Somewhat Strongly Agree 52. The sex offender registry makes life very difficult for sex offenders living in the community. Strongly Disagree Disagree Somewhat Unsure Agree Somewhat Strongly Agree 53. Children are safer if the locations of se x offenders are known. Strongly Disagree Disagree Somewhat Unsure Agree Somewhat Strongly Agree 243

PAGE 244

54. It is justified when individuals retaliate against sex offenders. Strongly Disagree Disagree Somewhat Unsure Agree Somewhat Strongly Agree 55. Individuals who retaliate against sex offenders should be subject to legal action. Strongly Disagree Disagree Somewhat Unsure Agree Somewhat Strongly Agree 56. Sex offenders should be released into the community after their prison sentences. Strongly Disagree Disagree Somewhat Unsure Agree Somewhat Strongly Agree 57. After a certain number of years a registered sex offender should be able to be removed from the registry. Str ongly Disagree Disagree Somewhat Unsure Agree Somewhat Strongly Agree 58. It will be too harsh to make sex offenders wear a special kind of marker, at all times on their person, which identifies them as a sex offender. Strongly Disa gree Disagree Somewhat Unsure Agree Somewhat Strongly Agree 59. Having to register on the Florida Sex Offender Registry constitutes cruel and unusual punishment. Strongly Disagree Disagree Somewhat Unsure Agree Somewhat Strongly Agree 60. I would support legislative action, which calls for sex offenders to be supervised under electronic monitoring/GPS tracking for the remainder of their lives while living in communities. Strongly Disagree Disagree Somewh at Unsure Agree Somewhat Strongly Agree 244

PAGE 245

Research Instrument (Qualitative) The following survey will ask you questions about your perceptions of sex offenders who may be living in communities near you. You will be asked questions about how much sex offenders concern you in terms of safety, how you feel about the sex offender reg istry and your reaction to sex offender registry laws and your knowledge about the sex offender registry. As stated in the informed consent, you may stop taking the survey at any time and you may skip any question that you are not willing to answer. 1. What is your gender? Male Female 2. How old are you? 18 20 21 23 24 26 27 or older 3. What is your race? Black White Other 4. Do you identify yourself as Hispanic? No Yes 5. What is your class standing? Freshman Sophomore Junior Senior This portion of the interview asks you to give in depth answers concerning your knowledge of the sex offender registry rules and of sex offenders that may be living in your neighborhood. Please provide an answer that best describes these issues. 6. How many times have you looked at the Florida Department of Law Enforcement, Florida Sex Offender Registry website? 0 times 1 time 2 times 3 times 4 or more times 7. What was your main reason for accessing or not accessing the website? Free Response 8. Do you feel as though the knowledge you have concerning the sex offender registry is accurate? Not very accurate A bit accurate Not sure Somewhat accurate Very Accurate 9. What informational outlets do you feel helped to form your knowledge base concerning the registry? Free Response 10. What information have you learned concerning the registry (feel free to give examples of your knowledge concerning the registr y)? Free Response 11. Based on your knowledge, what do you feel is the difference between a sexual offender and a sexual predator? Free Response 245

PAGE 246

This portion of the interview asks you to give in depth answers concerning your personal worri es, concerns and feelings regarding the registry and the sex offenders that may be living near you. Please provide an answer that best describes these issues. 12. How worried are you about sex offenders living nearby your home? Not really worried A little worried Somewhat worried Worried a lot 13. What are some of your reasons for being worried (or for not being worried) about sex offenders living nearby your home? Free Response 14. How angry are you that sex offenders are allowed to live in the community? Not really angry A little angry Somewhat angry Very Angry 15. What are some of your reasons for being angry (or for not being angry) about sex offenders being permitted to live in the community? Free Response 16. How mu ch are community members in agreement about the risk that sex offenders pose while living in the community? Not at all A little Somewhat A lot 17. What are some of your reasons for feeling that community members are in agreement (or are not in agr eement) about the risk that sex offenders pose while living in the community? Free Response 18. Legislators work quickly to get necessary registry laws passed to further keep track of sex offenders? Strongly Disagree Disagree Unsure Agree Stro ngly Agree 19. What are some of your reasons for feeling that legislators work fast enough (or don't work fast enough) to get necessary registry laws passed to further keep track of sex offenders? Free Response This portion of the interview asks you to provide in depth answers concerning your reaction to sexual offenses in general and about the laws associated with the registry. Please provide an answer that best describes these issues. 20. Do you support the use of the publicly available Flor ida Sex Offender Registry? Not really A little Somewhat A lot 21. What are some of your reasons for supporting the use of (or not supporting the use of) the publicly available Florida Sex Offender Registry? Free Response 22. Do you feel that the registry needs to be reformed? Why or why not? Free Respons e 246

PAGE 247

APPENDIX B PILOT STUDY CODING S HEET Quantitative Question Response Option Coding 1. What is your gender? Male 0 Female 1 2. How old are you? 18 20 1 21 23 2 24 26 3 27 or older 4 3. What is your race? Black 0 White 1 Other 2 4. Do you identify yourself as Hispanic? No 0 Yes 1 5. What is your class standing? Freshman 1 Sophomore 2 Junior 3 Senior 4 6. Please indicate if you have been a victim of any of the following crimes. Property Crime (such as theft or breaking & entering) 1 Drug Crime 2 Personal Crime: Non Sexual (such as robbery, assault or battery) 3 Personal Crime: Sexual (such as rape, sexual assault or molestation) 4 Other 5 Registry Access 7. How many times have you looked at the Florida Department of Law Enforcement, Florida Sex Offender Registry website? 0 times 0 1 time 1 2 times 2 3 times 3 247

PAGE 248

4 or more times 4 8. How many times have you looked at the website to search the areas surrounding your home for sex offenders living nearby? 0 times 0 1 time 1 2 times 2 3 times 3 4 or more times 4 9. How many sex offenders are you personally acquainted with? 0 offenders 0 1 offender 1 2 offenders 2 3 offenders 3 4 or more offenders 4 Registry Knowledge 10. Do you know that there are two types of registered sex offenders in Florida sexual offenders and sexual predators? No 0 Yes 1 This Isn't True 2 11. Do you know that registered sex offenders are required to live at least 1,000 feet from a school zone, park or bus stop? No 0 Yes 1 This Isn't True 2 12. Do you know that some sex offenders are required to register for life? No 0 Yes 1 This Isn't True 2 13. Do you know that not all sex offenders are required to be on some sort of electronic monitoring/GPS tracking device? No 0 Yes 1 This Isn't True 2 14. Where did you learn the most information regarding sex offenders and the sex offender registry? Television 1 Newspaper 2 Magazines 3 Radio 4 Internet 5 School 6 Friends/Family 7 Other 8 15. Everyone who has ever been convicted of a sexual offense is required to register on the Florida Sex Offender Registry. Very True 1 Somewhat True 2 Unsure 3 Somewhat False 4 Very False 5 248

PAGE 249

16. The definition of a sexual predator is: "Repeat sexual offenders, sexual offenders who use physical violence, and sexual offenders who prey on children are sexual predators who present an extreme threat to the public safety. Sexual offenders are extremely likely to use physical violence and to repeat their offenses, and most sexual offenders commit many offenses, have many more victims than are ever reported, and are prosecuted for only a fraction of their crimes." Very True 1 Somewhat True 2 Unsure 3 Somewhat False 4 Very False 5 17. Individuals convicted of their very first sexual crime can be classified as a sexual predator on the Florida Sex Offender Registry. Very True 1 Somewhat True 2 Unsure 3 Somewhat False 4 Very False 5 18. Juvenile offenders can be required to register on the Florida Sex Offender Registry if convicted of a sexual offense. Very True 1 Somewhat True 2 Unsure 3 Somewhat False 4 Very False 5 19. After serving their prison sentences, sex offenders can be incarcerated indefinitely through the process called "Civil Commitment." Very True 1 Somewhat True 2 Unsure 3 Somewhat False 4 Very False 5 20. In Florida, there are more male sex offenders registered than female sex offenders. Very True 1 Somewhat True 2 Unsure 3 Somewhat False 4 Very False 5 21. Female sex offenders make up roughly one fourth of the total amount of registered sex offenders. Very True 1 Somewhat True 2 Unsure 3 Somewhat False 4 Very False 5 22. The Amber Alert System is named after the color amber, at no time was the system designed in memory of a child. Very True 1 Somewhat True 2 Unsure 3 249

PAGE 250

Somewhat False 4 Very False 5 23. Sex offenders have one of the highest recidivism or re offending rates of any offenders. Very True 1 Somewhat True 2 Unsure 3 Somewhat False 4 Very False 5 Concern 24. Are you worried about sex offenders living nearby your home? Not Really 1 A Little 2 Somewhat 3 A Lot 4 25. Are you worried that you personally may be a victim of a sexual offense? Not Really 1 A Little 2 Somewhat 3 A Lot 4 26. Are you concerned about sex offenders being on the college campus? Not Really 1 A Little 2 Somewhat 3 A Lot 4 27. Are you worried that if sex offenders are living in the community, then more sexual offenses will occur? Not Really 1 A Little 2 Somewhat 3 A Lot 4 Hostility 28. Are you angry that sex offenders are allowed to live in the community? Not Really 1 A Little 2 Somewhat 3 A Lot 4 29. Do you feel any resentment over the fact that some of your neighbors may be sex offenders? Not Really 1 A Little 2 Somewhat 3 A Lot 4 30. Do you feel any anger towards the criminal justice system for releasing sex offenders from jails and prisons? Not Really 1 A Little 2 Somewhat 3 A Lot 4 31. Are you angry that sex offenders may be Not Really 1 250

PAGE 251

working at businesses where you may frequently shop or visit? A Little 2 Somewhat 3 A Lot 4 Consensus 32. Do you feel that a majority of community members are in agreement about the risk that sex offenders pose? Not Really 1 A Little 2 Somewhat 3 A Lot 4 33. Do you feel that many community members feel that changes must be made in the supervision of sex offenders? Not Really 1 A Little 2 Somewhat 3 A Lot 4 34. Do you feel that community members in general feel threatened by sex offenders as a group? Not Really 1 A Little 2 Somewhat 3 A Lot 4 35. Do you feel that a majority of community members are in agreement that children are at risk of being sexually victimized? Not Really 1 A Little 2 Somewhat 3 A Lot 4 36. Do you feel that many parents feel that sex offenders are too dangerous to be living in the community? Not Really 1 A Little 2 Somewhat 3 A Lot 4 Disproportionality 37. Do you feel that the current state of the sex offender registry is too harsh? Not Really 1 A Little 2 Somewhat 3 A Lot 4 38. Do you feel that the sex offender registry laws should be stricter? Not Really 1 A Little 2 Somewhat 3 A Lot 4 39. Do you feel that keeping sex offenders on electronic monitoring/GPS tracking for more than 5 years without a break is too severe a punishment. Not Really 1 A Little 2 Somewhat 3 A Lot 4 40. Do you feel that sex offenders should report to Not Really 1 251

PAGE 252

law enforcement more than the required two times per year? A Little 2 Somewhat 3 A Lot 4 41. Do you feel that the media overreacts in their reporting of sex offenses when they occur in a community? Not Really 1 A Little 2 Somewhat 3 A Lot 4 Volatility 42. Do you feel that law enforcement reacts quickly enough when a sexual offense takes place? Not Really 1 A Little 2 Somewhat 3 A Lot 4 43. Do you feel that legislators work fast enough to get necessary registry laws passed to further keep track of sex offenders? Not Really 1 A Little 2 Somewhat 3 A Lot 4 44. Do you feel that the media reports sex offenses cases too quickly before all of the facts are gathered? Not Really 1 A Little 2 Somewhat 3 A Lot 4 45. Do you feel that the quick response of the media makes communities safer because people are made aware of the sex offense? Not Really 1 A Little 2 Somewhat 3 A Lot 4 46. Do you feel that police are too slow to catch sex offenders when sexual offenses take place? Not Really 1 A Little 2 Somewhat 3 A Lot 4 Registry Support 47. Do you support the use of the publicly available Florida Sex Offender Registry? Not Really 1 A Little 2 Somewhat 3 A Lot 4 48. How often have you heard of sex offenders being hurt or killed by community members as a result of their sex offender status? 0 times 0 1 time 1 2 times 2 3 times 3 4 or more times 4 252

PAGE 253

49. How often have you seen fliers, notes or posters advertising the presence of a sexual offender who is living within your neighborhood? 0 times 0 1 time 1 2 times 2 3 times 3 4 or more times 4 50. Reforms should be made to the Florida Sex Offender Registry. Strongly Disagree 1 Disagree Somewhat 2 Unsure 3 Agree Somewhat 4 Strongly Agree 5 51. The registry is effective in reducing sex offender re offending. Strongly Disagree 1 Disagree Somewhat 2 Unsure 3 Agree Somewhat 4 Strongly Agree 5 52. The sex offender registry makes life very difficult for sex offenders living in the community. Strongly Disagree 1 Disagree Somewhat 2 Unsure 3 Agree Somewhat 4 Strongly Agree 5 53. Children are safer if the locations of sex offenders are known. Strongly Disagree 1 Disagree Somewhat 2 Unsure 3 Agree Somewhat 4 Strongly Agree 5 54. It is justified when individuals retaliate against sex offenders. Strongly Disagree 1 Disagree Somewhat 2 Unsure 3 Agree Somewhat 4 Strongly Agree 5 55. Individuals who retaliate against sex offenders should be subject to legal action. Strongly Disagree 1 Disagree Somewhat 2 253

PAGE 254

Unsure 3 Agree Somewhat 4 Strongly Agree 5 56. Sex offenders should be released into the community after their prison sentences. Strongly Disagree 1 Disagree Somewhat 2 Unsure 3 Agree Somewhat 4 Strongly Agree 5 57. After a certain number of years a registered sex offender should be able to be removed from the registry. Strongly Disagree 1 Disagree Somewhat 2 Unsure 3 Agree Somewhat 4 Strongly Agree 5 58. It will be too harsh to make sex offenders wear a special kind of marker, at all times on their person, which identifies them as a sex offender. Strongly Disagree 1 Disagree Somewhat 2 Unsure 3 Agree Somewhat 4 Strongly Agree 5 59. Having to register on the Florida Sex Offender Registry constitutes cruel and unusual punishment. Strongly Disagree 1 Disagree Somewhat 2 Unsure 3 Agree Somewhat 4 Strongly Agree 5 60. I would support legislative action, which calls for sex offenders to be supervised under electronic monitoring/GPS tracking for the remainder of their lives while living in communities. Strongly Disagree 1 Disagree Somewhat 2 Unsure 3 Agree Somewhat 4 Strongly Agree 5 254

PAGE 255

APPENDIX C PILOT STUDY FREQUENC Y AND REGRESSION TAB LES Participant Demographics (n = 667 ) Demographic Variable Response Choices Frequency Percentage Gender 0 = Male n = 217 32.5 % 1 = Female n = 450 67.5 % Age 1 = 18 20 n = 459 68.8 % 2 = 21 23 n = 183 27.4 % 3 = 24 26 n = 15 2.3 % 4 = 27 and older n = 10 1.5 % Race 1 = Black n = 97 14.6 % 2 = White n = 445 66.7 % 3 = Other n = 125 18.7 % Identify as Hispanic 0 = No n = 518 77.7 % 1 = Yes n = 149 22.3 % Academic Class Standing 1 = Freshman n = 14 8 22.2 % 2 = Sophomore n = 132 19.8 % 3 = Junior n = 233 34.9 % 4 = Senior n = 151 22.6 % Previous Victimization of a Crime 1 = Property Crime n = 157 23.5 % 2 = Drug Crime n = 9 1.3 % 3 = Personal Crime, Non Sexual n = 49 7.4 % 4 = Personal Crime, Sexual n = 42 6.4 % 5 = Other n = 40 6.0 % 6 = Never Been a Victim n = 370 55.4 % 255

PAGE 256

Registry Knowledge (n = 667 ) Demographic Variable Response Choices Frequency Percentage #7 How many times have you looked at the Florida Department of Law Enforcement, Florida Sex Offender Registry website? #8 How many times have you looked at the registry website to search the areas surrounding your home for sex offenders living nearby? 0 = 0 times 1 = 1 time 2 = 2 times 3 = 3 times 4 = 4 or more times 0 = 0 times n = 342 n = 135 n = 115 n = 23 n = 51 n = 336 51.3% 20.4% 17.2% 3.4% 7.6% 51.2 % 1 = 1 time n = 156 23.4 % 2 = 2 times n = 99 14.8 % 3 = 3 times n = 29 4.3 % 4 = 4 or more times n = 44 6.1 % #9 How many sex offenders are you personally acquainted with? 0 = 0 offenders 1 = 1 offender 2 = 2 offenders 3 = 3 offenders 4 = 4 or more offenders n = 599 n = 58 n = 9 n = 0 n = 1 89.9% 8.7% 1.3% 0.0% 0.1% #10 Do you know that there are two types of registered sex offenders in Florida sexual offenders and sexual predators? 0 = No n = 329 49.3 % 1 = Yes n = 314 47.1 % 2 = This isn't true n = 24 3.6 % #11 Do you know that registered sex offenders are required to live at least 1,000 feet from a school zone, park or bus stop? 0 = No n = 206 30.9% 1 = Yes n = 430 64.5 % 2 = This isn't true n = 31 4.6 % #12 Do you know that some sex offenders are required to register for life? 0 = No n = 118 17.7 % 1 = Yes n = 534 80.1% 2 = This isn't true n = 15 2.2 % #13 Do you know that not all sex offenders are required to be on some sort of electronic monitoring/GPS tracking device? 0 = No 1 = Yes 2 = This isn't true n = 627 n = 36 n = 4 94.0% 5.4% 0.6% #14 Where did you learn the most information regarding sex offenders and the sex offender registry? 1 = Television n = 239 35 .8% 2 = Newspaper n = 19 2.8 % 3 = Magazines n = 1 0.1 % 4 = Radio n = 2 0.3 % 5 = Internet n = 78 11.7 % 6 = School n = 204 30.6 % 7 = Friends/Family n = 103 15.4 % 8 = Other n = 21 3.1 % 256

PAGE 257

Levels of Accuracy for Registry Knowledge (n =667 ) Registry Knowledge Measure Response Choice Frequency Percentage #15 Everyone who has ever been convicted of a sexual offense is required to register on the Florida Sex Offender Registry. 1 = Very True 2 = Somewhat True 3 = Unsure 4 = Somewhat False 5 = Very False n = 272 n = 197 n = 143 n = 36 n = 19 40.8% 29.5% 21.4% 5.4% 2.8% #16 The definition of a sexual predator is: Repeat sexual offenders, sexual offenders who use physical violence, and sexual offenders who prey on children are sexual predators who present an extreme threat to the public safety. Sexual offenders are extremely likely to use physical violence and to repeat their offenses, and most sexual offenders commit many offenses, have many more victims than are ever reported, and are prosecuted for only a fraction of their crimes. 1 = Very True 2 = Somewhat True 3 = Unsure 4 = Somewhat False 5 = Very False n = 169 n = 253 n = 180 n = 49 n = 16 25.3% 37.9% 27.0% 7.3% 2.4% #17 Individuals convicted of their very first sexual crime can be classified as a sexual predator on the Florida Sex Offender Registry. 1 = Very True n = 141 21.1 % 2 = Somewhat True n = 168 25.2 % 3 = Unsure n = 273 40.9 % 4 = Somewhat False n = 60 9.0 % 5 = Very False n = 25 3.7 % #18 Juvenile offenders can be required to register on the Florida Sex Offender Registry, if convicted of a sexual offense. 1 = Very True n = 162 24.3 % 2 = Somewhat True n = 181 27.1 % 3 = Unsure n = 254 38.1 % 4 = Somewhat False n = 49 7 .3% 5 = Very False n = 21 3.1 % #19 After serving their prison sentences, sex offenders can be incarcerated indefinitely through the process called "Civil Commitment." 1 = Very True n = 50 7.5 % 2 = Somewhat True n = 83 12.4 % 3 = Unsure n = 439 65.8 % 4 = Somewhat False n = 89 13.3 % 5 = Very False n = 6 0.9 % #20 In Florida, there are more male sex offenders registered than female sex offenders. 1 = Very True 2 = Somewhat True 3 = Unsure 4 = Somewhat False 5 = Very False n = 337 n = 197 n = 125 n = 2 n = 6 50.5% 29.5% 18.7% 0.3% 0.9% 257

PAGE 258

#21 Female sex offenders make up roughly one fourth of the total amount of registered sex offenders. 1 = Very True 2 = Somewhat True 3 = Unsure 4 = Somewhat False 5 = Very False n = 48 n = 165 n = 363 n = 84 n = 7 7.2% 24.7% 54.4% 12.6% 1.0% #22 The Amber Alert System is named after the color amber, at no time was the system designed in the memory of a child. 1 = Very True 2 = Somewhat True 3 = Unsure 4 = Somewhat False 5 = Very False n = 29 n = 20 n = 143 n = 403 n = 72 4.3% 3.0% 21.4% 60.4% 10.8% #23 Sex offenders have one of the highest recidivism rates of any offenders. 1 = Very True n = 136 20.4 % 2 = Somewhat True n = 218 32.7 % 3 = Unsure n = 145 36.7 % 4 = Somewhat False n = 59 8.8 % 5 = Very False n = 9 1.3 % 258

PAGE 259

Concern or Fear Rel ated to Sexual Offenders (n = 66 7) Concern/Fear Measure Response Option Frequency Percentage #24 Are you worried about sex offenders living nearby your home? 1 = Not Really n = 236 35.4 % 2 = A Little n = 236 35.4 % 3 = Somewhat n = 129 19.3 % 4 = A Lot n = 66 9.9 % #25 Are you worried that you personally may be a victim of a sexual offense? 1 = Not Really n = 346 51.9 % 2 = A Little n = 144 21.6 % 3 = Somewhat n = 109 16.3 % 4 = A Lot n = 68 10.2 % #26 Are you concerned about sex offenders being on the college campus? 1 = Not Really 2 = A Little 3 = Somewhat 4 = A Lot n = 141 n = 203 n = 179 n = 144 21.1% 30.4% 26.8% 21.6% #27 Are you worried that if sex offenders are living in the community, then more sexual offenses will occur? 1 = Not Really n = 136 20.4 % 2 = A Little n = 230 34.5 % 3 = Somewhat n = 203 30.4% 4 = A Lot n = 98 14.7 % Hostility or Anger Rel ated to Sexual Offenders (n = 66 7) Hostility Measure Response Option Frequency Percentage #28 Are you angry that sex offenders are allowed to live in the community? 1 = Not Really n = 291 43.6 % 2 = A Little n = 194 29.1 % 3 = Somewhat n = 116 17.4 % 4 = A Lot n = 66 9.9 % #29 Do you feel any resentment over the fact that some of your neighbors may be sex offenders? 1 = Not Really n = 245 36.7 % 2 = A Little n = 228 34.2 % 3 = Somewhat n = 133 19.9 % 4 = A Lot n = 61 9.1 % #30 Do you feel any anger towards the criminal justice system for releasing sex offenders from jails and prisons? 1 = Not Really 2 = A Little 3 = Somewhat 4 = A Lot n = 279 n = 199 n = 137 n = 52 41.8% 29.8% 20.5% 7.8% #31 Are you angry that sex offenders may be working at businesses where you may frequently shop or visit? 1 = Not Really n = 287 43.0 % 2 = A Little n = 2 09 31.3 % 3 = Somewhat n = 114 17.1 % 4 = A Lot n = 57 8.5 % 259

PAGE 260

Community Consensus Rel ated to Sexual Offenders (n = 66 7) Consensus Measure Response Option Frequency Percentage #32 Do you feel that a majority of community members are in agreement about the risk that sex offenders pose? 1 = Not Really n = 92 13.8 % 2 = A Little n = 184 27.6 % 3 = Somewhat n = 247 37.0 % 4 = A Lot n = 144 21.6 % #33 Do you feel that many community members feel that changes must be made in the supervision of sex offenders? 1 = Not Really n = 88 13.2 % 2 = A Little n = 2 24 32.1 % 3 = Somewhat n = 248 37.2 % 4 = A Lot n = 117 17.5 % #34 Do you feel that community members in general feel threatened by sex offenders as a group? 1 = Not Really 2 = A Little 3 = Somewhat 4 = A Lot n = 62 n = 170 n = 259 n = 176 9.3 % 25.5 % 38.8 % 26.4 % #35 Do you feel that a majority of community members are in agreement that children are at risk of being sexually victimized? 1 = Not Really n = 44 6.6 % 2 = A Little n = 121 18.1 % 3 = Somewhat n = 217 32.5 % 4 = A Lot n = 285 42.7 % #36. Do you feel that many parents feel that sex offenders are too dangerous to be living in the community? 1 = Not Really 2 = A Little 3 = Somewhat 4 = A Lot n = 25 n = 86 n = 185 n = 371 3.7% 12.9% 27.7% 55.6% 260

PAGE 261

Disproportionality of Response Rel ated to Sexual Offenders (n = 66 7) Disproportionality Measure Response Option Frequency Percentage #37. Do you feel that the current state of the sex offender registry is too harsh? 1 = Not Really n = 444 66 .6 % 2 = A Little n = 1 42 23.1 % 3 = Somewhat n = 64 9.6 % 4 = A Lot n = 17 2.5 % #38. Do you feel that the sex offender registry laws should be stricter? 1 = Not Really n = 193 28.9 % 2 = A Little n = 2 32 34.8 % 3 = Somewhat n = 1 60 24.0 % 4 = A Lot n = 82 12.3 % #39. Do you feel that keeping sex offenders on electronic monitoring/GPS tracking for more than 5 years without a break is too severe a punishment. 1 = Not Really 2 = A Little 3 = Somewhat 4 = A Lot n = 401 n = 135 n = 93 n = 38 60.1 % 20.2 % 12.9 % 5.7 % #40. Do you feel that sex offenders should report to law enforcement more than the required two times per year? 1 = Not Really n = 84 12.6 % 2 = A Little n = 151 22.6 % 3 = Somewhat n = 1 80 27.0 % 4 = A Lot n = 252 37.8 % #41. Do you feel that the media overreacts in their reporting of sex offenses when they occur in a community? 1 = Not Really 2 = A Little 3 = Somewhat 4 = A Lot n = 316 n = 171 n = 106 n = 74 47.4% 25.6% 15.9% 11.1% 261

PAGE 262

Volatility of Response Related to Sexual Offenders (n = 517) #42 Do you feel that law enforcement reacts quickly enough when a sexual offense takes place? 1 = Not Really n = 168 25.2 % 2 = A Little n = 208 31.2 % 3 = Somewhat n = 238 35.7 % 4 = A Lot n = 53 7.9% #43 Do you feel that legislatures work fast enough to get necessary registry laws passed to further keep track of sex offenders? 1 = Not Really n = 246 36.9 % 2 = A Little n = 237 35.5 % 3 = Somewhat n = 162 24.3 % 4 = A Lot n = 22 3.3 % #44 Do you feel that the media reports sex offenses too quickly before all of the facts are gathered? 1 = Not Really n = 135 20.2 % 2 = A Little n = 200 30.0 % 3 = Somewhat n = 192 28.8 % 4 = A Lot n = 140 21.0 % #45. Do you feel that the quick response of the media makes communities safer because people are made aware of the sex offense? 1 = Not Really 2 = A Little 3 = Somewhat 4 = A Lot n = 214 n = 200 n = 176 n = 77 32.1% 30.0% 26.4% 11.5% #46 Do you feel that police are too slow to catch sex offenders when sexual offenses take place? 1 = Not Really n = 173 25.9 % 2 = A Little n = 280 42.0 % 3 = Somewhat n = 156 23.4 % 4 = A Lot n = 58 8.7 % 262

PAGE 263

Support for the Registry (n = 517) #47 Do you support the use of the publicly available Florida Sex Offender Registry? 1 = Not Really n = 32 4.8 % 2 = A Little n = 75 11.2 % 3 = Somewhat n = 140 21.0 % 4 = A Lot n = 420 63.0 % #48 How often have you heard of sex offenders being hurt or killed by community members as a result of their sex offender status? 0 = 0 Times 1 = 1 Time 2 = 2 Times 3 = 3 Times 4 = 4 or more Times n = 586 n = 48 n = 0 n = 11 n = 22 87.9% 7.2% 0.0% 1.6% 3.3% #49 How often have you seen fliers, notes or posters advertising the presence of a sexual offender who is living within your neighborhood? 0 = 0 Times 1 = 1 Time 2 = 2 Times 3 = 3 Times 4 = 4 or more Times n = 540 n = 64 n = 0 n = 17 n = 46 81.0% 9.6% 0.0% 2.5% 6.9% #50 Reforms should be made to the Florida Sex Offender Registry. 1 = Strongly Disagree n = 19 2.8 % 2 = Disagree Somewhat n = 81 12.1 % 3 = Unsure n = 332 49.8 % 4 = Agree Somewhat n = 193 28.9 % 5 = Strongly Agree n = 42 6 .3% #51 The registry is effective in reducing sex offender re offending. 1 = Strongly Disagree n = 38 5.7 % 2 = Disagree Somewhat n = 135 20.2 % 3 = Unsure n = 293 43.9 % 4 = Agree Somewhat n = 177 26.5 % 5 = Strongly Agree n = 24 3.6 % #52 The sex offender registry makes life very difficult for sex offenders living in the community. 1 = Strongly Disagree n = 19 2.8 % 2 = Disagree Somewhat n = 88 13.2 % 3 = Unsure n = 111 16.6 % 4 = Agree Somewhat n = 330 49.5 % 5 = Strongly Agree n = 119 17.8 % #53 Children are safer if the locations of sex offenders are known. 1 = Strongly Disagree n = 32 4.8 % 2 = Disagree Somewhat n = 72 10.8 % 3 = Unsure n = 89 13.3 % 4 = Agree Somewhat n = 291 43.6 % 5 = Strongly Agree n = 183 27.4 % # 54 It is justified when individuals retaliate against sex offenders. 1 = Strongly Disagree 2 = Disagree Somewhat 3 = Unsure 4 = Agree Somewhat 5 = Strongly Agree n = 243 n = 215 n = 99 n = 73 n = 37 36.4% 32.2% 14.8% 10.9% 5.5% #55 Individuals who retaliate against sex offenders should be subject to legal action. 1 = Strongly Disagree 2 = Disagree Somewhat 3 = Unsure 4 = Agree Somewhat 5 = Strongly Agree n = 29 n = 49 n = 82 n = 252 4.3% 7.3% 12.3% 37.8% 263

PAGE 264

n = 255 38.2% #56 Sex offenders should be released into the community after their prison sentences. 1 = Strongly Disagree 2 = Disagree Somewhat 3 = Unsure 4 = Agree Somewhat 5 = Strongly Agree n = 67 n = 181 n = 179 n = 300 n = 40 10.0% 27.1% 26.8% 30.0% 6.0% #57 After a certain number of years a registered sex offender should be able to be removed from the registry. 1 = Strongly Disagree 2 = Disagree Somewhat 3 = Unsure 4 = Agree Somewhat 5 = Strongly Agree n = 221 n = 178 n = 108 n = 134 n = 26 33.1% 26.7% 16.2% 20.1% 3.9% #58 It would be too harsh to make sex offenders wear a special kind of marker, at all times on their person, which identifies them as a sex offender. 1 = Strongly Disagree n = 60 9.0 % 2 = Disagree Somewhat n = 107 16.0 % 3 = Unsure n = 93 13.9 % 4 = Agree Somewhat n = 199 29.8 % 5 = Strongly Agree n = 208 31.2 % #59 Having to register on the Florida Sex Offender Registry constitutes cruel and unusual punishment. 1 = Strongly Disagree n = 380 57.0 % 2 = Disagree Somewhat n = 179 26.8 % 3 = Unsure n = 78 11.7 % 4 = Agree Somewhat n = 23 3.4 % 5 = Strongly Agree n = 7 1.0 % #60 I would support legislative action, which calls for sex offenders to be supervised under electronic monitoring/GPS tracking for the remainder of their lives while living in communities. 1 = Strongly Disagree 2 = Disagree Somewhat 3 = Unsure 4 = Agree Somewhat 5 = Strongly Agree n = 112 n = 168 n = 156 n = 130 n = 101 16.8% 25.2% 23.4% 19.5% 15.1% 264

PAGE 265

Bivariate Correlations For Registry Knowledge Q15 Everyone Required to Register Q16 Sex Predator Definition Q17 First Time Predators Q18 Juvenile SOs Q19 Civil Commitment Q20 Males Outnumber Females Q21 Number of Females Q22 Amber Alert Q23 Reduce Recidivism Q15 Everyone Required to Register Pearson Correlation Sig. (2 tailed) N 1 667 Q16 Sex Predator Definition Pearson Correlation Sig. (2 tailed) N .106** .006 667 1 667 Q17 First Time Predators Pearson Correlation Sig. (2 tailed) N .052 .176 667 .050 .201 667 1 667 Q18 Juvenile SOs Pearson Correlation Sig. (2 tailed) N .098* .012 667 .008 .836 667 .190** .000 667 1 667 Q19 Civil Commitmen t Pearson Correlation Sig. (2 tailed) N .036 .356 667 .148** .000 667 .078* .043 667 0.94* .015 667 1 667 Q20 Males Outnumber Females Pearson Correlation Sig. (2 tailed) N .070 .071 667 .083* .033 667 .119** .002 667 0.38 .323 667 056 .152 667 1 667 Q21 Number of Females Pearson Correlation Sig. (2 tailed) N .009 .818 667 .020 .613 667 .052 .179 667 .069 .073 667 .187** .000 667 055 .158 667 1 667 Q22 Amber Alert Pearson Correlation Sig. (2 tailed) N .055 .006 667 .070 .073 667 064 .100 667 .115** .003 667 .120 ** .002 667 077 .046 667 053 .169 667 1 667 Q23 Reduce Recidivism Pearson Correlation Sig. (2 tailed) N .026 .508 667 .182** .000 667 .024 .544 667 .057 .139 667 .065 .094 667 .162 ** .000 667 087* .025 667 035 .360 667 1 667 265

PAGE 266

Bivariate Correlations For Moral Panic Element of Concern Q24 SOs Live Nearby Q25 Being Victimized Q26 SOs on College Campus Q27 Increased Offending Q24 SOs Live Nearby Pearson Correlation Sig. (2 tailed) N 1 667 Q25 Being Victimized Pearson Correlation Sig. (2 tailed) N 439** .000 667 1 667 Q26 SOs on College Campus Pearson Correlation Sig. (2 tailed) N 564** .000 667 .507** .000 667 1 667 Q27 Increased Offending Pearson Correlation Sig. (2 tailed) N 510** .0 00 667 375** .000 667 651 ** .000 667 1 667 266

PAGE 267

Bivariate Correlations For Moral Panic Element of Hostility Q28 Angry that SOs Live in Nearby Q29 Neighbors that are SOs Q30 CJS Released SOs Q31 SOs work at businesses you frequent Q28 Angry that SOs Live in Nearby Pearson Correlation Sig. (2 tailed) N 1 667 Q29 Neighbors that are SOs Pearson Correlation Sig. (2 tailed) N 665** .000 667 1 667 Q30 CJS Released SOs Pearson Correlation Sig. (2 tailed) N 564** .000 667 .491** .000 667 1 667 Q31 SOs work at businesses you frequent Pearson Correlation Sig. (2 tailed) N 671** .0 00 667 626** .000 667 630 ** .000 667 1 667 267

PAGE 268

Bivariate Correlations For Moral Panic Element of Consensus Q 32 Risk that SOs Pose Q 33 Change in Supervision Q 34 Threatened by SOs Q 35 Children At Risk Q36 Parents see the danger from SOs Q 32 Risk that SOs Pose Pearson Correlation Sig. (2 tailed) N 1 667 Q 33 Change in Supervision Pearson Correlation Sig. (2 tailed) N 357** .000 667 1 667 Q 34 Threatened by SOs Pearson Correlation Sig. (2 tailed) N 294** .000 667 .487** .000 667 1 667 Q 35 Children At Risk Pearson Correlation Sig. (2 tailed) N 326** .0 00 667 409** .000 667 514 ** .000 667 1 667 Q36 Parents see the danger from SOs Pearson Correlation Sig. (2 tailed) N 265** .0 00 667 344** .0 00 667 483** .0 00 667 501** .0 00 667 1 667 268

PAGE 269

Bivariate Correlations For Moral Panic Element of Disproportionality Q 37 Registry is too Harsh Q 38 Registry needs to be stricter Q 39 GPS for 5 years too severe Q 40 Increased reporting to law enforcement Q41 Media overreacts in reporting Q 37 Registry is too Harsh Pearson Correlation Sig. (2 tailed) N 1 667 Q 38 Registry needs to be stricter Pearson Correlation Sig. (2 tailed) N .263** .000 667 1 667 Q 39 GPS for 5 years too severe Pearson Correlation Sig. (2 tailed) N 468** .000 667 .247 ** .000 667 1 667 Q 40 Increased reporting to law enforcement Pearson Correlation Sig. (2 tailed) N .178** .0 00 667 445** .000 667 .204 ** .000 667 1 667 Q41 Media overreacts in reporting Pearson Correlation Sig. (2 tailed) N .288** .0 00 667 .159** .0 00 667 360** .0 00 667 .140 ** .0 00 667 1 667 269

PAGE 270

Bivariate Correlations For Moral Panic Element of Volatility Q 42 Law enforcement react quickly Q 43 Legislators work fast enough Q 44 Media reports too quickly Q 45 Media reporting makes us safer Q46 Police are too slow Q 42 Law enforcement react quickly Pearson Correlation Sig. (2 tailed) N 1 667 Q 43 Legislators work fast enough Pearson Correlation Sig. (2 tailed) N .447** .000 667 1 667 Q 44 Media reports too quickly Pearson Correlation Sig. (2 tailed) N 091* .019 667 .078* .045 667 1 667 Q 45 Media reporting makes us safer Pearson Correlation Sig. (2 tailed) N .097* .0 12 667 107** .006 667 .370 ** .000 667 1 667 Q46 Police are too slow Pearson Correlation Sig. (2 tailed) N .290** .0 00 667 .164** .0 00 667 .061 117 667 .152** .0 00 667 1 667 270

PAGE 271

OLS Regression Predicting Moral Panic Element of Concern Variable B SE Registry Knowledge 0.022 .011 .006 Gender 438*** 064 259 Age .064 .061 .048 Race .083 .056 .060 Ethnicity .213*** .077 .1 12 Class Standing .162*** .033 .218 Property Crime Victimization .087 .069 .047 Drug Crime Victimization .031 .130 .009 Non Sexual Violent Crime Victimization .045 .038 .044 Sexual Violent Crime Victimization .017 .030 .021 Other Victimization .000 .025 .001 Constant 2.201 .329 F Statistic 9.186 *** R Square .135 *** p<.001 **p < .05 *p < .01 271

PAGE 272

OLS Regression Predicting Moral Panic Element of Hostility Variable B SE Registry Knowledge .004 .01 2 .0 13 Gender .151 .070 .086 Age .133 .066 .096 Race .013 .061 .009 Ethnicity .052 .083 .026 Class Standing .153 *** .036 .199 Property Crime Victimization .099 .076 .051 Drug Crime Victimization .190 .141 .054 Non Sexual Violent Crime Victimization .038 .042 .036 Sexual Violent Crime Victimization .028 .033 .034 Other Victimization .006 .027 .008 Constant 1.923 .357 F Statistic 2.793** R Square .045 *** p<.001 **p < .05 *p < .01 272

PAGE 273

OLS Regression Predicting Moral Panic Element of Consensus Variable B SE Registry Knowledge .023** .009 .094 Gender .043 .056 .031 Age .054 .053 .049 Race .023 .049 .020 Ethnicity .030 .067 .019 Class Standing .073 ** .029 .119 Property Crime Victimization .016 .061 .010 Drug Crime Victimization .079 .114 .028 Non Sexual Violent Crime Victimization .014 .034 .017 Sexual Violent Crime Victimization .033 .027 .050 Other Victimization .029 .022 .052 Constant 2.373 .289 F Statistic 1.591 R Square .026 *** p<.001 **p < .05 *p < .01 273

PAGE 274

OLS Regression Predicting Moral Panic Element of Disproportionality Variable B SE Registry Knowledge .013 ** .007 .077 Gender .052 .039 .054 Age .000 .037 .000 Race .049 .034 .062 Ethnicity .019 .046 .017 Class Standing .029 .020 .068 Property Crime Victimization .039 .042 .037 Drug Crime Victimization .142 .079 .073 Non Sexual Violent Crime Victimization .024 .023 .041 Sexual Violent Crime Victimization .013 .018 .027 Other Victimization .020 .015 .052 Constant 1.713 .199 F Statistic 1.502 R Square .025 *** p<.001 **p < .05 *p < .01 274

PAGE 275

OLS Regression Predicting Moral Panic Element of Volatility Variable B SE Registry Knowledge .00 9 .006 .060 Gender .000 .037 .000 Age .032 .035 .044 Race .024 .032 .032 Ethnicity .018 .044 .017 Class Standing .051 ** .019 .128 Property Crime Victimization .033 .040 .033 Drug Crime Victimization .086 .074 .047 Non Sexual Violent Crime Victimization .012 .022 .022 Sexual Violent Crime Victimization .003 .017 .007 Other Victimization .024 .014 .068 Constant 2.002 .187 F Statistic 1.470 R Square .0 2 4 *** p<.001 **p < .05 *p < .01 275

PAGE 276

OLS Regression Predicting Registry Support Variable B SE Registry Knowledge .010 .006 .066 Gender .074 ** .034 .086 Age .039 .032 .057 Race .031 .030 .045 Ethnicity .030 .041 .031 Class Standing .002 .018 .006 Property Crime Victimization .051 .037 .054 Drug Crime Victimization .163 ** .069 .094 Non Sexual Violent Crime Victimization .012 .020 .024 Sexual Violent Crime Victimization .024 .016 .060 Other Victimization .001 .013 .002 Constant 3.144 .175 F Statistic 1.813** R Square .030 *** p<.001 **p < .05 *p < .01 276

PAGE 277

Bivariate Correlations for Registry Knowledge (RK) and Control Variables (Full Study). RK Stereot ype Parent # Kids Gend er Age Ethnic ity Ed. Level Marital Status Income Level SES Geo. Region Pop Size RK Correlation Sig. (2 tailed) N 1 877 Stereotype Correlation Sig. (2 tailed) N .062 .066 877 1 877 Parental Correlation Sig. (2 tailed) N .016 .636 877 0.59 .080 877 1 877 # Kids Correlation Sig. (2 tailed) N .040 .234 877 .016 .646 877 .475*** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .044 .188 877 .003 .923 877 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .048 .157 877 .224*** .000 877 .395*** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .001 .978 877 .025 .468 877 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .003 .936 877 .136*** .000 877 .014 .681 877 .016 .626 877 .041 .228 877 .110*** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .007 .830 877 .052 .126 877 .270*** .000 877 .148*** .000 877 .024 .476 877 .079* .019 877 .051 .124 877 .134*** .000 877 1 877 Income Level Correlation Sig. (2 tailed) N .054 .107 877 .005 .879 877 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238*** .000 877 .220*** .000 877 1 877 SES Correlation Sig. (2 tailed) N .025 .467 877 .108*** .001 877 .004 .902 877 .009 .789 877 .007 .844 877 .038 .260 877 .001 .986 877 .297*** .000 877 .224** .000 877 .417*** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .007 .840 877 .050 .138 877 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 877 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .023 .493 877 .006 .854 877 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 877 .116*** .001 877 .127*** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 277

PAGE 278

Bivariate Correlations for Registry Access and Control Variables (Full Study). Registry Access Parent # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Registry Access Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .073 .030 877 1 877 # Kids Correlation Sig. (2 tailed) N .082* .015 877 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .041 .221 877 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .011 .751 877 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .044 .197 877 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .072* .032 877 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .084* .013 875 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .064 .059 877 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .074* .029 877 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .072* .032 877 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .073* .031 877 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 278

PAGE 279

Bivariate Correlations for Fear of Personal Victimization and Control Variables (Full Study). Fear of Personal Victim Paren t # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Fear of Personal Victim. Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .059 .082 877 1 877 # Kids Correlation Sig. (2 tailed) N .028 .403 877 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .068* .043 877 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .128** .000 877 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .037 .270 877 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .026 .442 877 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .029 .393 875 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .007 .833 877 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .066 .051 877 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .018 .602 877 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .013 .709 877 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 279

PAGE 280

Bivariate Correlations for Fear of Victimization of Children and Control Variables (Full Study). Fear of Victim of Children Parent # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Fear of Victim of Children Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .210** .000 877 1 877 # Kids Correlation Sig. (2 tailed) N .123** .000 877 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .062 .065 877 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .089** .008 877 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .043 .204 877 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .022 .509 877 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .040 .239 875 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .053 .114 877 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .020 .559 877 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .016 .643 877 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .001 .969 877 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 280

PAGE 281

Bivariate Correlations for Moral Panic Concern (Offender) and Control Variables (Full Study). Concern (OFF) Parent # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Concern (OFF) Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .195** .000 439 1 877 # Kids Correlation Sig. (2 tailed) N .144** .003 439 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .103* .030 439 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .072 .135 439 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .076 .110 439 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .028 .555 439 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .029 .548 438 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .028 .553 439 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .025 .607 439 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .049 .302 439 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .040 .402 439 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 281

PAGE 282

Bivariate Correlations for Moral Panic Concern (Predator) and Control Variables (Full Study). Concern (PRED) Parent # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Concern (PRED) Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .192** .000 413 1 877 # Kids Correlation Sig. (2 tailed) N .115* .019 413 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .187** .000 413 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .085 .086 413 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .045 .366 413 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .076 .124 413 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .056 .254 412 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .016 .742 413 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .086 .080 413 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .067 .174 413 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .009 .861 413 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 282

PAGE 283

Bivariate Correlations for Moral Panic Hostility (Offender) and Control Variables (Full Study). Hostility (OFF) Parent # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Hostility (OFF) Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .196** .000 439 1 877 # Kids Correlation Sig. (2 tailed) N .113* .018 439 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .040 .408 439 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .054 .255 439 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .038 .423 439 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .025 .605 439 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .024 .615 438 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .059 .218 439 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .033 .495 439 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .056 .245 439 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .053 .267 439 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 283

PAGE 284

Bivariate Correlations for Moral Panic Hostility (Predator) and Control Variables (Full Study). Hostility (PRED) Parent # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Hostility (PRED) Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .194** .000 413 1 877 # Kids Correlation Sig. (2 tailed) N .115* .019 413 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .201** .000 413 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .029 .556 413 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .064 .193 413 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .077 .118 413 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .075 .127 412 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .022 .656 413 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .064 .192 413 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .026 .598 413 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .048 .333 413 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 284

PAGE 285

Bivariate Correlations for Moral Panic Consensus (Offender) and Control Variables (Full Study). Consens us (OFF) Parent # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Consensu s (OFF) Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .037 .442 439 1 877 # Kids Correlation Sig. (2 tailed) N .037 .440 439 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .026 .587 439 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .131** .006 439 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .085 .076 439 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .029 .550 439 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .060 .210 438 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .005 .913 439 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .078 .103 439 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .079 .096 439 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .020 .680 439 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 285

PAGE 286

Bivariate Correlations for Moral Panic Consensus (Predator) and Control Variables (Full Study). Consen sus (PRED) Parent # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Consensu s (PRED) Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .031 .524 413 1 877 # Kids Correlation Sig. (2 tailed) N .108* .028 413 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .211** .000 413 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .104* .035 413 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .069 .164 413 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .028 .569 413 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .031 .533 412 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .042 .394 413 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .010 .841 413 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .000 .992 413 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .016 .742 413 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 286

PAGE 287

Bivariate Correlations for Moral Panic Disproportionality (Offender) and Control Variables (Full Study). Dispropo rtionality (OFF) Paren t # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Disproporti onality (OFF) Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .154** .001 439 1 877 # Kids Correlation Sig. (2 tailed) N .101* .035 439 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .040 .408 439 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .210** .000 439 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .030 .526 439 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .124** .009 439 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .011 .823 438 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .059 .221 439 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .071 .135 439 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .083 .083 439 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .117* .014 439 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 287

PAGE 288

Bivariate Correlations for Moral Panic Disproportionality (Predator) and Control Variables (Full Study). Hostility (PRED) Parent # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Hostility (PRED) Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .156** .001 413 1 877 # Kids Correlation Sig. (2 tailed) N .118* .016 413 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .292** .000 413 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .151** .002 413 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .080 .102 413 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .043 .382 413 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .111* .024 412 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .019 .698 413 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .052 .289 413 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .064 .192 413 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .015 .755 413 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 288

PAGE 289

Bivariate Correlations for Moral Panic Volatility (Offender) and Control Variables (Full Study). Volatility (OFF) Parent # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Volatility (OFF) Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .044 .359 439 1 877 # Kids Correlation Sig. (2 tailed) N .009 .843 439 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .054 .259 439 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .094* .048 439 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .085 .077 439 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .100* .037 439 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .035 .466 438 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .006 .899 439 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .006 .903 439 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .042 .380 439 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .046 .332 439 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 289

PAGE 290

Bivariate Correlations for Moral Panic Volatility (Predator) and Control Variables (Full Study). Volatility (PRED) Parent # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Volatility (PRED) Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .002 .968 413 1 877 # Kids Correlation Sig. (2 tailed) N .010 .843 413 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .072 .146 413 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .073 .140 413 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .015 .764 413 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .045 .360 413 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .020 .683 412 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .066 .178 413 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .001 .980 413 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .022 .655 413 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .033 .499 413 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 290

PAGE 291

Bivariate Correlations for CATSO Social Isolation and Control Variables (Full Study). Social Isolation Parent # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Social Isolation Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .029 .393 877 1 877 # Kids Correlation Sig. (2 tailed) N .024 .470 877 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .003 .940 877 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .157** .000 877 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .050 .137 877 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .088** .009 877 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .084* .013 875 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .024 .471 877 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .041 .229 877 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .038 .260 877 .004 .902 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .002 .962 877 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 291

PAGE 292

Bivariate Correlations for CATSO Capacity to Change and Control Variables (Full Study). Capacity to Change Paren t # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Capacity to Change Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .176** .000 877 1 877 # Kids Correlation Sig. (2 tailed) N .100** .003 877 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .054 .112 877 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .013 .710 877 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .015 .649 877 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .066* .050 877 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .125** .000 875 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .006 .855 877 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .048 .155 877 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .018 .592 877 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .064 .057 877 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 292

PAGE 293

Bivariate Correlations for CATSO Severity/Dangerousness and Control Variables (Full Study). Severity /Danger Parent # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Severity/D anger Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .080* .018 877 1 877 # Kids Correlation Sig. (2 tailed) N .028 .408 877 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .045 .184 877 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .200** .000 877 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .002 .956 877 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .150** .000 877 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .064 .058 875 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .032 .343 877 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .081* .016 877 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .043 .203 877 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .012 .716 877 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 2 93

PAGE 294

Bivariate Correlations for CATSO Deviancy and Control Variables (Full Study). Deviancy Paren t # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Deviancy Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .106** .002 877 1 877 # Kids Correlation Sig. (2 tailed) N .088** .009 877 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .005 .891 877 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .051 .129 877 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .005 .891 877 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .061 .070 877 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .062 .067 875 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .024 .469 877 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .017 .624 877 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .019 .581 877 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .041 .228 877 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 294

PAGE 295

Bivariate Correlations for CATSO Total Index and Control Variables (Full Study). Total Index Parent # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Total Index Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .125** .000 877 1 877 # Kids Correlation Sig. (2 tailed) N .062 .067 877 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .028 .415 877 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .126** .000 877 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .024 .481 877 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .116** .001 877 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .107** .001 875 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .028 .405 877 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .059 .083 877 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .012 .720 877 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .030 .377 877 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 295

PAGE 296

. Bivariate Correlations for Registry Support (Offender) and Control Variables (Full Study). Registry Support (OFF) Parent # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Registry Support (OFF) Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .049 .309 438 1 877 # Kids Correlation Sig. (2 tailed) N .037 .444 438 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .016 .734 438 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .167** .000 438 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .039 .413 438 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .144** .003 438 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .022 .654 437 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .074 .120 438 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .044 .356 438 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .074 .122 438 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .019 .699 438 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 296

PAGE 297

Bivariate Correlations for Registry Support (Predator) and Control Variables (Full Study). Registry Support (PRED) Parent # Kids Gender Age Ethnicit y Ed. Level Marital Status Income Level SES Geo. Region Pop Size Registry Support (PRED) Correlation Sig. (2 tailed) N 1 877 Parent Correlation Sig. (2 tailed) N .026 .593 411 1 877 # Kids Correlation Sig. (2 tailed) N .003 .950 411 .475** .000 877 1 877 Gender Correlation Sig. (2 tailed) N .100* .042 411 .058 .086 877 .040 .232 877 1 877 Age Correlation Sig. (2 tailed) N .121* .014 411 .395** .000 877 .107** .002 877 .061 .071 877 1 877 Ethnicity Correlation Sig. (2 tailed) N .001 .986 411 .044 .197 877 .003 .938 877 .008 .812 877 .012 .731 877 1 877 Ed. Level Correlation Sig. (2 tailed) N .002 .972 411 .014 .681 877 .016 .626 877 .041 .228 877 .110** .001 877 .012 .723 877 1 877 Marital Status Correlation Sig. (2 tailed) N .045 .359 410 .270** .000 875 .148** .000 875 .024 .476 875 .079* .019 875 .051 .134 875 .134** .000 875 1 877 Income Level Correlation Sig. (2 tailed) N .098* .047 411 .032 .350 877 .009 .787 877 .026 .446 877 .022 .515 877 .008 .816 877 .238** .000 877 .220** .000 875 1 877 SES Correlation Sig. (2 tailed) N .048 .336 411 .004 .902 877 .009 .798 877 .007 .844 877 .038 .260 877 .001 .986 877 .297** .000 877 .224** .000 875 .417** .000 877 1 877 Geo. Region Correlation Sig. (2 tailed) N .067 .177 411 .026 .446 877 .021 .540 877 .032 .349 877 .059 .081 877 .016 .643 877 .060 .077 877 .015 .648 875 .041 .223 877 .059 .083 877 1 877 Pop. Size Correlation Sig. (2 tailed) N .035 .484 411 .077* .022 877 .040 .239 877 .036 .293 877 .086* .011 877 .022 .516 877 .095** .005 877 .075* .026 875 .116** .001 877 .127** .000 877 .090** .008 877 1 877 *The variables are abbreviated for the size of the table. 297

PAGE 298

Sex Offenders in the Community: How Much Do Parents Really Know About Sex Offenders? Informed Consent Dear Participant, Thank you so much for agreeing to participate in this assessment of community member perceptions of the sex offender registry and of sexual offenders. Your HONEST feedback is greatly appreciated. To start the survey questionnaire, click on the NEXT arrow at the bottom of the page. It will take 30 60 minutes to complete the survey. If you need any help, please feel free to contact Jennifer Klein (jklein87@ufl.edu) or Lonn Lanza Kaduce, at llkkll@ufl.edu. Please read this informed cons ent carefully before agreeing to participate in this study. You must read this informed consent befo re participating in this study. Purpose of the Research Study: The purpose is to examine how much participants know about the Florida Sex Offender Registr y, sex offenders in general and how individuals perceive registered sex offenders. What You will be Asked to do in the Research Study: If you decide to participate in this survey, you will be asked to answer questions about your perceptions of the sex offender registry and about those individuals required to register. As a participant, all identifiable information will remain confidential. This means that no information (name, home address, email/IP address or other identifying information) will be ab le to be traced back to you. Your responses will be kept anonymous, meaning that none of the answers you provide are linked back to you as an individual. The survey will include questions about your level of knowledge in regards to sex offender laws, c haracteristics of sex offenders and your perceptions surrounding this subject. There will also be questions about your demographic characteristics. Your individual answers will not be shared with anyone. Please do not provide specific information a bout yourself beyond what is being asked in the participant demographic questions. The questions asking about personal information at the beginning of the survey are not specific enough to be able to identify you compared to any other participant. When y ou are finished with the survey, the researcher will send your gift card to you in the mail. The information you provide will be stored in a secure computer (this computer is guarded with a firewall and spyware). None of the answers provided will be able to be traced back to you as a participant. Time Required: It will take between 30 minutes and an hour to answer the questions, depending on your pace. Confidentiality: All of your answers will be kept anonymous, meaning that no one will be able to link your answers back to you as an individual. Your personal information such as your name, home address or email/IP address, will be kept private and confidential to the exte nt provided by law; it will not be stored with your answers and will 298

PAGE 299

be retained separately, only long enough to send a reminder email and a gift card to your home. The results of the study will present patterns of how everyone answered. The report will not focus on any one person's answers. Voluntary Participation and Right to Withdraw From the Study: By participating, you will only benefit monetarily. Participants will be sent a $10 VISA gift card after completing the study. The gift card wi ll be sent to the address provided in the contact information. The risks are minimal, but you may find that you are uncomfortable answering some of the questions. All of the questions are opinion based in nature and none of your answers will expose you to any legal ramifications. You are free to talk about any issues that might be raised for you by the survey to any and all counseling services that are available to you. You do not have to answer any questions that you do not want to answer and you can stop taking the survey at any time. No one will be upset or angry if you decide not to participate or if you stop participating at any time for any reason. Whom to Contact if you Have Questions About the Study: Jennifer Klein or Dr. Lonn Lanza Ka duce, Department of Sociology, Criminology and Law, 3219 Turlington Hall, PO Box 117330, Gainesville, Florida 32611 7330; Telephone: (352) 392 0265; Email: jklein87@ufl.edu Whom to Contact About Your Rights as a Research Participant in the Study: UFI RB Office, Box 112250, University of Florida, Gainesville, FL 32611 2250; Telephone: (352) 392 0433 or irb2@ufl.edu Agreement: By providing and submitting your contact information you will consent to participate in this study. By completing the su rvey, the assumption will be made that you have read the informed consent and are fine with the researcher using your answers for academic purposes only. Your gift card will be sent to the home address that you have provided. What is your contact informa tion? (only used to send out your gift card) First Name (1) Last Name (2) Address 1 (3) Address 2 (4) City (5) State (6) Zip Code (7) I have read the consent form and am willing to participate in this study. Yes (1) No (2) If Yes Is Not Selected, Then Skip To You have chosen not to participate in... Thank you for agreeing to participate in the study... Below is the link to be taken to the survey. 1) Click the link to open the survey in a new browser2) PLEASE PRESS THE 299

PAGE 300

NEXT BUTTON (>>>) to close out this survey and so that your name is recorded. https://ufl.qualtrics.com/SE/?SID=SV_bEPkqiUnEwTCikB YOU WILL NEED TO PUT IN THE PASSWORD: gogators You have chosen not to participate in the survey. If this is in error please press the back button and give consent. Then you will be able to proceed to this survey. If this was not in error thank you for your time! 300

PAGE 301

"Sex Offenders in the Community: How Much Do Parents Really Know About Sex Offenders?" The following survey will ask you questions about your pe rceptions of sex offenders who may be living nearby you and your family. You will be asked questions about how much sex offenders concern you in terms of safety, how you feel about the sex offender registry and your reaction to sex offender registry laws. What is your gender? Male Female How old are you? Drop down scale of 18 100 years old What is your race? Native American/Alaskan Asian Native Hawaiian or other Pacific Is lander Black/African American White Other Do you identify yourself as Hispanic? No Yes Please answer the following questions about your children. 1 Child 2 Children 3 Children 4 or more Children How many children do you have? How many of your children are currently enrolled in a schoo l located in Alachua County? How old is your first child? 0 4 5 9 10 14 15 18 19 and older How old is your second child? 0 4 5 9 10 14 15 18 19 and older How old is your third child? 0 4 5 9 10 14 301

PAGE 302

15 18 19 and older How old is your fourth child? 0 4 5 9 10 14 15 18 19 and older Please answer the following questions about your children. 0 Children 1 Child 2 Children 3 Children 4 or more Children How many of your childr en are male? How ma ny of your children are female? What is your marital status? Single Married Divorced Widowed Other ____________________ What is your current family income level? $0 $50,000 $50,0001 $100,000 $100,001 $150,000 $150,001 $200,000 $200,001 and up What socio economic class do you most identify with? Lower Class Middle Class Upper Middle Class Upper Class 302

PAGE 303

Please answer the following two questions about victimization (select all that apply). Property Crime (such as theft, or breaking and entering ) Drug Crime Personal Crime: Non Sexual (such as robbery, assault, or battery) Personal Crime: Sexual (such as rape, sexual assault or molestation) Other Please indicate if you have been a victim of any of the following crimes. Please indicate if any of your close family members (children, siblings, parents, etc.) or close friends have been the victim of any of the following crimes. You have indicated that you have been a victim of a crime that falls into the "other" category. What is that crime? (Please just name the indicated offense WITHOUT describing it in detail). You have indicated that you have been a victim of a crime that falls into the "other" category. What is that crime? (Please just name the indicated o ffense WITHOUT describing it in detail). Have you ever looked at the Florida Department of Law Enforcement, Florida Sex Offender Registry website? No Yes Every year the Alachua County Sheriff's Office prints a booklet of registered sexual offenders a nd predators that is distributed in the Gainesville Sun newspaper. Have you ever looked at this booklet? No Yes Do you support the use of the publicly available Florida Sex Offender Registry? Definitely not Probably not Probably yes Definitely yes 303

PAGE 304

How strict do you think the current laws regarding the Florida Sex Offender Registry are? Way Too Lax Lax Just Right Strict Way Too Strict Have you ever searched the website or newspaper booklet for sex offenders living in the areas surrounding your home? No Yes How many sex offenders would you estimate are living nearby your home? Drop down scale of 0 to 100 or more Are you afraid that a sexual offender could victimize your child while they are at school (during the course of the day or during extra curricular activities)? Definitely not Probably not Probably yes Definitely yes How many sex offenders would you estimate are living around your child's school? Drop down scale of 0 to 100 or more Have you interacted with anyone living near you that you know to be a sex offender? No Yes (2) If so, how many individuals are you are aware of? Drop down scale of 1 to 100 or more Have you interacted with anyone not living near you that you know to be a sex offender? No Y es If so, how many individuals are you aware of? Drop down scale of 1 to 100 or more The following questions test your knowledge of the Florida Sex Offender Registry and the state laws regarding the restrictions placed on sexual offenders. Please answer as accurately as possible. 304

PAGE 305

Very True Somewhat True Unsure Somewhat False Very False There is only one type of registered sex offenders in Florida. Registered sex offenders are required to live at least 1,000 feet from a school zone, park or bus stop. Some sex offenders are required to register for life. Juvenile offenders, who are 14 years old at the time of the offense, can be placed on the registry if convicted. All sex offenders are required to be on some sort of electronic monitoring/GPS tracking device at all times. Sex offenders have very high rates of reoffending. The Amber Alert System is named after a child named Amber, it has nothing to do with the color amber. In Florida, there are more male sex offenders registered than female sex 305

PAGE 306

offenders. Individuals convicted of their very first sexual crime can be classified as a sexual predator on the Florida Sex Offender Registry. After serving their prison sentences, sex offenders can be incarcerated indefinitely though a process called Civil C ommitment. Where have you learned ANY information regarding sex offenders and the sex offender registry? Television Newspaper Magazines Radio Internet School Friends/Family Other ____________________ Where do you learn the MOST information regarding sex offenders and the sex offender registry? Television Newspaper Magazines Radio Internet School Friends/Family Other ____________________ 306

PAGE 307

The following questions ask about your personal worries and concerns regarding sexual offenders. Definitely not Probably not Probably yes Definitely yes Are you worried about sex offenders living nearby your home? Are you worried that you personally may become a victim of a sexual offense? Are you concerned that your children may be at risk of becoming victims of a sexual offense? Are you worried about your children being at risk of being approached by a sexual offender? Are you worried that as sex offenders continue to live in the community, then more sex offenses will occur? 307

PAGE 308

The following questions ask about your possible feelings of anger concerning sexual offenders. Definitely not Probably not Probably yes Definitely yes Are you angry that sex offenders are allowed to live in the community? Do you feel any resentment over the fact that some of your neighbors may be sex offenders? Do you feel any anger towards the criminal justice system for releasing sex offenders from jails and prisons? Are you angry that sex offenders may be working at businesses that you may frequently shop at or visit? Are you angry that your chil dren might come into contact with a sex offender? 308

PAGE 309

The following questions ask about community consensus and unity regarding sexual offenders. Definitely not Probably not Probably yes Definitely yes Do you think that a majority of parents like yourself are in agreement about the risk that sex offenders pose? Do you think that many parents feel that changes must be make in the supervision of sex offenders? Do you think that parents in general feel threatened by sex offenders as a group? Do you think that a majority of parents are in agreement that their children are at risk of being sexually victimized? Do you think that many parents feel that sex offenders are too dangerous to be living in the community? 309

PAGE 310

The following questions ask about a possible disproportionate community response regarding sexual offenders. Definitely not Probably not Probably yes Definitely yes Do you feel that the current state of the sex offender registry is too harsh? Do you feel that the sex offender registry laws should be stricter? Do you think that keeping sex offenders on electronic monitoring/GPS tracking for more then 5 years without a break is too severe a punishment? Do you think that sex offenders should report to law enforcement more then th e required two times per yea r? Do you think that the media overreacts in their reporting of sex offenses when they occur in a community? 310

PAGE 311

The following questions ask about a possible volatile community reaction in response to sexual offenders. Definitely not Probably not Probably yes Definitely yes Do you think that law enforcement reacts quickly enough when a sexual offense takes place? Do you think that legislators work fast enough to get necessary registry laws passed to further keep tract of sex offenders? Is it possible that the media reports sex offense cases too quickly before all of the facts are gathered? Do you think that the quick response of the media makes communities safer because people are made aware of the sex offense? Do you think that police are too slow to catch sex offenders when sex offenses take place? 311

PAGE 312

This portion of the survey asks you about your reaction to sexual offenses in general and about the laws associated with the registry. Please select the most appropriate answer that describes your point of view on these issues. How many times in the past 12 m onths have you heard of sex offenders being hurt or killed by community members as a result of their sex offender status? Never 1 Time 2 Times 3 Times 4 Times 5 Times 6 Times 7 Times 8 Times 9 Times 10 or More Times How many times in the past 12 months have you seen fliers, notes, or posters advertising the presence of a sexual offender who is living within your neighborhood? Never 1 Time 2 Times 3 Times 4 Times 5 Times 6 Times 7 Times 8 Times 9 Times 10 or More Tim es 312

PAGE 313

The following questions address your personal opinion regarding the Florida Sex Offender Registry and sex offenders in general. Thinking about Florida specifically, please answer as honestly as possible. Strongly Disagree Disagree Unsure Agre e Strongly Agree Reforms should be made to the sex offender registry. The sex offender registry is effective in reducing sex offender re offen d ing. The sex offender registry makes life very difficult for sex offenders living in the community. Children are safer if the locations of sex offenders are known. It is justified when individuals retaliate against sex offenders. Individuals who retaliate against sex offenders should be subject to legal action. Sex offenders should be released into the community after their prison sentences. After a certain number of years, a registered sex 313

PAGE 314

offender should be able t o be removed from the sex offender registry. It will be too harsh to make sex offenders wear a special kind of marker, at all times on their person, which identifies them as a sex offender. Having to register on the sex offender registry constitutes cruel and unusual punishment. I would support legislative action which calls for sex offenders to be supervised under electronic monitoring/GPS tracking for the remainder of their lives w hile living in communities. I support residency restrictions that prevent sex offenders from living too closely to schools, playgrounds and other areas where children frequently gather. 314

PAGE 315

Below are 18 statements about sex offenders and sex offenses. Please select the corresponding number from the rating scale given below for the answer that best describes the way you feel or what you believe. Most of the statements below are difficult to prove or verify in an absolute sense, and many are spe cifically about your opinion based on what you may have heard, read, or learned; thus, we are less interested in the "right" or "wrong" answers, and more interested in your beliefs and opinions regarding sex offenders. Even if you have no general knowledge about the issue, please provide an answer to each question. Strongly Disagree Disagree Probably Disagree Probably Agree Agree Strongly Agree With support and therapy, someone who committed a sexual offense can learn to change their behavior. People who commit sex offenses should lose their civil rights (e.g. voting and privacy). People who commit sex offenses want to have sex more often than the average person. Male sex offenders should be punished more severely than female sex offenders. A lot of sex offenders use their victims to create pornography. Sex offenders prefer to stay 315

PAGE 316

home alone rather than to be around lots of people. Most sex offenders do not have close friends. Sex offenders have difficulty making friends even if they try real hard. The prison sentences sex offenders receive are much too long when compared to the sentence lengths for other crimes. Sex offenders have high rates of sexual activity. Trying to rehabilitate a sex offender is a waste of time. Sex offenders should wear tracking devices so their location can be pinpointed at any time. Only a few sex offenders are dangerous. 316

PAGE 317

Most sex offenders are unmarried men. Someone who uses emotional control when committing a sex offense is not as bad as someone who uses physical control when committing a sex offense. Most sex offenders keep to themselves. A sex offense committed against someone the perpetrato r knows is less serious than a sex offense committed against a stranger. Convicted sex offenders should never be released from prison. Do you have any recommendations to make towards the registry, sex offender supervision, community living or job regulations that would make a difference to you in terms of eliminating the threat that sex offenders pose in your life? You have now reached the end of the survey!! Thank you for your participation and for the responses you have provided. 317

PAGE 318

Sex Offenders in the Community: How Much Do Community Members Really Know About Sex Offenders? Informed Consent Dear Participant, Thank you so much for agreeing to participate in this assessment of community member perceptions of the sex offender registry and of sexual offenders. Your HONEST feedback is greatly appreciated. To start the survey questionnaire, click on the NEXT arrow at the bottom of the page. It will take 30 60 minutes to complete the survey. If you need any help, plea se feel free to contact Jennifer Klein (jklein87@ufl.edu) or Lonn Lanza Kaduce, at llkkll@ufl.edu. Please read this informed consent carefully before agreeing to participate in this study. You must read this informed consent befo re participating in thi s study. Purpose of the Research Study: The purpose is to examine how much participants know about the Florida Sex Offender Registry, sex offenders in general and how individuals perceive registered sex offenders. What You will be Asked to do in the Re search Study: If you decide to participate in this survey, you will be asked to answer questions about your perceptions of the sex offender registry and about those individuals required to register. As a participant, all identifiable information will rem ain confidential. This means that no information (name, home address, email/IP address or other identifying information) will be able to be traced back to you. Your responses will be kept anonymous, meaning that none of the answers you provide are link ed back to you as an individual. The survey will include questions about your level of knowledge in regards to sex offender laws, characteristics of sex offenders and your perceptions surrounding this subject. There will also be questions about your demo graphic characteristics. Your individual answers will not be shared with anyone. Please do not provide specific information about yourself beyond what is being asked in the participant demographic questions. The questions asking about personal info rmation at the beginning of the survey are not specific enough to be able to identify you compared to any other participant. When you are finished with the survey, the researcher will send your gift card to you in the mail. The information you provide wi ll be stored in a secure computer (this computer is guarded with a firewall and spyware). None of the answers provided will be able to be traced back to you as a participant. Time Required: It will take between 30 minutes and an hour to answer the questions, depending on your pace. Confidentiality: All of your answers will be kept anonymous, meaning that no one will be able to link your answers back to you as an individual. Your personal information such as your name, home address or e mail/IP address, will be kept private and 318

PAGE 319

confidential to the extent provided by law; it will not be stored with your answers and will be retained separately, only long enough to send a reminder email and a gift card to your home. The results of the study will present patterns of how everyone answered. The report will not focus on any one person's answers. Voluntary Participation and Right to Withdraw From the Study: By participating, you will only benefit monetarily. Participants will be sent a $10 VISA gift card after completing the study. The gift card will be sent to the address provided in the contact information. The risks are minimal, but you may find that you are uncomfortable answering some of the questions. All of the questions are opinion based in nature and none of your answers will expose you to any legal ramifications. You are free to talk about any issues that might be raised for you by the survey to any and all counseling services that are available to you. You do not have to answer any questions that you do not want to answer and you can stop taking the survey at any time. No one will be upset or angry if you decide not to participate or if you stop participating at any time for any reason. Whom to Contact if you Have Questions About the Study: Jennifer Klein or Dr. Lonn Lanza Kaduce, Department of Sociology, Criminology and Law, 3219 Turlington Hall, PO Box 117330, Gainesville, Florida 32611 7330; Telephone: (352) 392 0265; Email: jklein87@ufl.edu Whom to Contact About Your Rights as a Research Participant in the Study: UFIRB Office, Box 112250, University of Florida, Gainesville, FL 32611 2250; Telephone: (352) 392 0433 or irb2@uf l.edu Agreement: By providing and submitting your contact information you will consent to participate in this study. By completing the survey, the assumption will be made that you have read the informed consent and are fine with the researcher usi ng your answers for academic purposes only. Your gift card will be sent to the home address that you have provided. What is your contact information? (only used to send out your gift card) First Name (1) Last Name (2) Address 1 (3) Address 2 (4) City (5) State (6) Zip Code (7) I have read the consent form and am willing to participate in this study. Yes No If Yes Is Not Selected, Then Skip To You have chosen not to participate in... 319

PAGE 320

Thank you for agreeing to participate in the study... Below is the li nk to be taken to the survey. 1) Click the link to open the survey in a new browser2) PLEASE PRESS THE NEXT BUTTON (>>>) to close out this survey and so that your name is recorded. https://ufl.qualtrics.com/SE/?SID=SV_bEPkqiUnEwTCikB YOU WILL NEED TO PUT IN THE PASSWORD: gogators You have chosen not to participate in the survey. If this is in error please press the back button and give consent. Then you will be able to proceed to this survey. If this was not in error thank you for your time! 320

PAGE 321

"Sex Offenders in the Community: How Much Do Community Members Really Know About Sex Offenders?" The following survey will ask you questions about your perceptions of sex offenders who may be living nearby you and your family. You will be asked questions about how much sex offenders concern you in terms of safety, how you feel about the sex offender registry and your reaction to sex offender registry laws. What is your gender? Male Female How old are you? Drop down scale o f 18 100 years old What is your race? Native American/Alaskan Asian Native Hawaiian or other Pacific Is lander Black/African American White Other Do you identify yourself as Hispanic? No Yes Please answer the following questions about your children. 1 Child 2 Children 3 Children 4 or more Children How many children do you have? How many of your children are currently enrolled in a schoo l located in Alachua County? How old is your first child? 0 4 5 9 10 14 15 18 19 and older How old is your second child? 0 4 5 9 10 14 15 18 19 and older How old is your third child? 0 4 5 9 321

PAGE 322

10 14 15 18 19 and older How old is your fourth child? 0 4 5 9 10 14 15 18 19 and older Please answer the following questions about your children. 0 Children 1 Child 2 Children 3 Children 4 or more Children How many of your childr en are male? How ma ny of your children are female? Do you intend to have children in the future? No Yes I already have children and I do not have plans to have more Are you a grandparent? No Yes How many grandchildren do you have? 1 2 3 4 or more 322

PAGE 323

Answer If Are you a grandparent? Yes Is Selected Please answer the following questions about your grandchildren. 0 Grandchildre n 1 Grandchild 2 Grandchildren 3 Grandchildren 4 or More Grandchildren How many of your grandchildren are currently enrolled in a school located in Alachua County? How many of your grandchildren are male? How many of your grandchildren are female? What is your marital status? Single Married Divorced Widowed Other ____________________ What is your current family income level? $0 $50,000 $50,0001 $100,000 $100,001 $150,000 $150,001 $200,000 $200,001 and up What socio economic class do you most identify with? Lower Class Middle Class Upper Middle Class Upper Class 323

PAGE 324

Please answer the following two questions about vic timization (select all that apply). Property Crime (such as theft, or breaking and entering) Drug Crime Personal Crime: Non Sexual (such as robbery, assault, or battery) Personal Crime: Sexual (such as rape, sexual assault or molestation) Other Please indicate if you have been a victim of any of the following crimes. Please indicate if any of your close family members (children, siblings, parents, etc.) or close friends have been the victim of any of the following crimes. You have indicated that you have been a victim of a crime that falls into the "other" category. What is that crime? (Please just name the indicated offense WITHOUT describing it in detail). You have indicated that you have been a victim of a crime that falls into the "other" category. What is that crime? (Please just name the indicated offense WITHOUT describing it in detail). Have you ever looked at the Florida Department of Law Enforcement, Florida Sex Offender Registry website? No Yes Every year the Alac hua County Sheriff's Office prints a booklet of registered sexual offenders and predators that is distributed in the Gainesville Sun newspaper. Have you ever looked at this booklet? No Yes Do you support the use of the publicly available Florida Sex Offender Registry? Definitely not Probably not Probably yes Definitely yes 324

PAGE 325

How strict do you think the current laws regarding the Florida Sex Offender Registry are? Way Too Lax Lax Just Right Strict Way Too Strict Have you ever searched the website or newspaper booklet for sex offenders living in the areas surrounding your home? No Yes How many sex offenders would you estimate are living nearby your home? Drop down scale of 0 to 100 or more Are you afraid that a sexual offender could victimize your child while they are at school (during the course of the day or during extra curricular activities)? Definitely not Probably not Probably yes Definitely yes How many sex offenders would you estimate are living around your child 's school? Drop down scale of 0 to 100 or more Have you interacted with anyone living near you that you know to be a sex offender? No Yes (2) If so, how many individuals are you are aware of? Drop down scale of 1 to 100 or more Have you interacted with anyone not living near you that you know to be a sex offender? No Yes If so, how many individuals are you aware of? Drop down scale of 1 to 100 or more The following questions test your knowledge of the Florida Sex Offender Registry an d the state laws regarding the restrictions placed on sexual offenders. Please answer as accurately as possible. 325

PAGE 326

Very True Somewhat True Unsure Somewhat False Very False There is only one type of registered sex offenders in Florida. Registered sex offenders are required to live at least 1,000 feet from a school zone, park or bus stop. Some sex offenders are required to register for life. Juvenile offenders, who are 14 years old at the time of the offense, can be placed on the registry if convicted. All sex offenders are required to be on some sort of electronic monitoring/GPS tracking device at all times. Sex offenders have very high rates of reoffending. The Amber Alert System is named after a child named Amber, it has nothing to do with the color amber. In Florida, there are more male sex offenders registered than female sex 326

PAGE 327

offenders. Individuals convicted of their very first sexual crime can be classified as a sexual predator on the Florida Sex Offender Registry. After serving their prison sentences, sex offenders can be incarcerated indefinitely though a process called Civil C ommitment. Where have you learned ANY information regarding sex offenders and the sex offender registry? Television Newspaper Magazines Radio Internet School Friends/Family Other ____________________ Where do you learn the MOST information regarding sex offenders and the sex offender registry? Television Newspaper Magazines Radio Internet School Friends/Family Other ____________________ 327

PAGE 328

The following questions ask about your personal worries and concerns regarding sexual offenders. Definitely not Probably not Probably yes Definitely yes Are you worried about sex offenders living nearby your home? Are you worried that you personally may become a victim of a sexual offense? Are you concerned that your children may be at risk of becoming victims of a sexual offense? Are you worried about your children being at risk of being approached by a sexual offender? Are you worried that as sex offenders continue to live in the community, then more sex offenses will occur? 328

PAGE 329

The following questions ask about your possible feelings of anger concerning sexual offenders. Definitely not Probably not Probably yes Definitely yes Are you angry that sex offenders are allowed to live in the community? Do you feel any resentment over the fact that some of your neighbors may be sex offenders? Do you feel any anger towards the criminal just ice system for releasing sex offenders from jails and prisons? Are you angry that sex offenders may be working at businesses that you may frequently shop at or visit? Are you angry that your children might come into contact with a sex offender? 329

PAGE 330

The following questions ask about community consensus and unity regarding sexual offenders. Definitely not Probably not Probably yes Definitely yes Do you think that a majority of parents like yourself are in agreement about the risk that sex offenders pose? Do you think that many parents feel that changes must be make in the supervision of sex offenders? Do you think that parents in general feel threatened by sex offenders as a group? Do you think that a majority of parents are in agreement that their children are at risk of being sexually victimized? Do you think that many parents feel that sex offenders are too dangerous to be living in the community? 330

PAGE 331

The following questions ask about a possible disproportionate community response regarding sexual offenders. Definitely not Probably not Probably yes Definitely yes Do you feel that the current state of the sex offender registry is too harsh? Do you feel that the sex offender registry laws should be stricter? Do you think that keeping sex offenders on electronic monitoring/GPS tracking for more then 5 years without a break is too severe a punishment? Do you think that sex offenders should report to law enforcement more then th e required two times per yea r? Do you think that the media overreacts in their reporting of sex offenses when they occur in a community? 331

PAGE 332

The following questions ask about a possible volatile community reaction in response to sexual offenders. Definitely not Probably not Probably yes Definitely yes Do you think that law enforcement reacts quickly enough when a sexual offense takes place? Do you think that legislators work fast enough to get necessary registry laws passed to further keep tract of sex offenders? Is it possible that the media reports sex offense cases too quickly before all of the facts are gathered? Do you think that the quick response of the media makes communities safer because people are made aware of the sex offense? Do you think that police are too slow to catch sex offenders when sex offenses take place? 332

PAGE 333

This portion of the survey asks you about your reaction to sexual offenses in general and about the laws associated with the registry. Please select the most appropriate answer that describes your point of view on these issues. How many times in the past 12 m onths have you heard of sex offenders being hurt or killed by community members as a result of their sex offender status? Never 1 Time 2 Times 3 Times 4 Times 5 Times 6 Times 7 Times 8 Times 9 Times 10 or More Times How many times in the past 12 months have you seen fliers, notes, or posters advertising the presence of a sexual offender who is living within your neighborhood? Never 1 Time 2 Times 3 Times 4 Times 5 Times 6 Times 7 Times 8 Times 9 Times 10 or More Times 333

PAGE 334

The following questions address your personal opinion regarding the Florida Sex Offender Registry and sex offenders in general. Thinking about Florida specifically, please answer as honestly as possible. Strongly Disagree Disagree Unsure Agree Strongly Agree Reforms should be made to the sex offender registry. The sex offender registry is effective in reducing sex offender re offen d ing. The sex offender registry makes life very difficult for sex offenders living in the community. Children are safer if the locations of sex offenders are known. It is justified when individuals retaliate against sex offenders. Individuals who retaliate against sex offenders should be subject to legal action. Sex offenders should be released into the community after their prison sentences. After a certain number of years, a registered sex 334

PAGE 335

offender should be able to be removed from the sex offender registry. It will be too harsh to make sex offenders wear a special kind of marker, at all times on their person, which identifies them as a sex offender. Having to register on the sex offender registry constitutes cruel and unusual punishment. I would support legislative action which calls for sex offenders to be supervised under electronic monitoring/GPS tracking for the remainder of their lives while living in communities. I support residency restrictions that prevent sex offenders from living too closely to schools, playgrounds and other areas where children frequently gather. 335

PAGE 336

Below are 18 statements about sex offenders and sex offenses. Please select the corresponding number from the rating scale given below for the answer that best describes the way you feel or what you believe. Most of the statements below are difficult to prove or verify in an absolute sense, and many are specifically about your opinion based on what you may have heard, read, or learned; thus, we are less interested in the "right" or "wrong" answers, and more interested in your beliefs and opinions regarding sex offenders. Even if you have no general knowledge about the issue, please pro vide an answer to each question. Strongly Disagree Disagree Probably Disagree Probably Agree Agree Strongly Agree With support and therapy, someone who committed a sexual offense can learn to change their behavior. People who commit sex offenses should lose their civil rights (e.g. voting and privacy). People who commit sex offenses want to have sex more often than the average person. Male sex offenders should be punished more severely than female s ex offenders. A lot of sex offenders use their victims to create pornography. Sex offenders prefer to stay 336

PAGE 337

home alone rather than to be around lots of people. Most sex offenders do not have close friends. Sex offenders have difficulty making friends even if they try real hard. The prison sentences sex offenders receive are much too long when compared to the sentence lengths for other crimes. Sex offenders have high rates of sexual activity. Trying to rehabilitate a sex offender is a waste of time. Sex offenders should wear tracking devices so their location can be pinpointed at any time. Only a few sex offenders are dangerous. 337

PAGE 338

Most sex offenders are unmarried men. Someone who uses emotional control when committing a sex offense is not as bad as someone who uses physical control when committing a sex offense. Most sex offenders keep to themselves. A sex offense committed against someone the perpetrator knows is less serious than a sex offense committed against a stranger. Convicted sex offenders should never be released from prison. Do you have any recommendations to make towards the registry, sex offender supervision, community living or job regulations that would make a difference to you in terms of eliminating the threat that sex offenders pose in your life? You have now reached the end of the survey!! Thank you for your participation and for the responses you have provided. 338

PAGE 339

LIST OF REFERENCES Adam Walsh Act of 2006, Public Law 10 9 248. (2006). Retrieved from: http://www.justice.gov/criminal/c eos/Adam%20Walsh.pdf Ackerman, A.R., Harris, A.J., Levenson, J.S. & Zgoba, K. (2011). Who are the people in your neighborhood? A descriptive analysis of individuals on public sex offender registries. International Journal of Law and Psychiatry, 34, 149 1 59. Adkins,G., Huff, D., & Stateberg, P. (2000). The Iowa sex offender registry and recidivism. Des Moines, IA: Iowa Department of Human Rights Division of Criminal and Juvenile Justice Planning and Statistical Analysis Center. Akers, R.L., LaGreca, A. J., Sellers, C. & Cochrane, J. (1987). Fear of crime and victimization among the elderly in different types of communities. Criminology, 25, 487 505. Alachua County Sheriff's Office, ACSO. (2012). Alachua County's Registered Sexual Offenders and Predators. Retrieved from: http://www.gainesville.com/assets/pdf/2012_Alachua_County_Sexual_Offenders. pdf Allport, G.W. (1954/1979). The Nature of Prejudice. Reading, MA: Addison Wesley. Angelides, S. (2003). Historicizing affect, psychoanalyzing histor y: Pedophilia and the discourse of child sexuality. Journal of Homosexuality Vol. 46, 1/2, pp. 79 109. Avrahamian, K. (1998). A critical perspective: Do Megan's Laws really shield children from sex predators? Journal of Juvenile Law 19 301 317. Back, S. & Lips, H.M. (1998). Child sexual abuse: Victim age, victim gender, and observer gender as factors contributing to attributions of responsibility. Child Abuse and Neglect Vol. 22, 1239 1252. Balow, K. & Conley, T.B. (2008). A report to the Mon tana Department of Corrections on community corrections professionals' attitudes towards sex offenders. University of Montana School of Social Work. Bates, A. & Metcalf, C. (2007). A psychometric comparison of Internet and non Internet sex offenders fro m a community treatment sample. Jou rnal of Sexual Aggression, 13, 11 20 Barlow, H.D. (1993). Introduction to Criminology (6 th edition). New York: Harper Collins. Baron Evans, A. & Noonan, S. (2006). The Adam Walsh Child Protection and Safety Ac t of 2006 -Part I. Retrieved from: http://www.fd.org/pdf_lib/Adam%20Walsh%20MemoPt%201.pdf 339

PAGE 340

Best, J. (1990). Threatened Children: Rhetoric and Concern about Child Victims. Chicago: University of Chicago Press. Braithwaite, J. (1989). Crime, shame, and reintegration Cambridge, UK: Cambridge University. Broussard, S., Wagner, W.G. & Kazelskis, R. (1991). Undergraduate students' perceptions of child sexual abuse: The impact of victim sex, offender sex, participant sex and victim response. Journal of Family Violence, 6 267 278. Bureau of Justice Assistance (2010). Background Information on the Act and Its Amendments. Retrieved from: http://www.ojp.usdoj.gov/BJA/what/2a2jwactbackground.html Burchfield, K.B. & Mingus, W. (2008). Not in my neighborhoo d: Assessing registered sex offenders' experiences with local social capital and social control. Criminal Justice and Behavior, 35 (3), 356 374. Calley, N.G. (2008). Juvenile sex offenders and sex offender legislation: Unintended consequences. Federal P robation 72 (3), 37 41. Caldwell, M. (2007). Sexual offense adjudication and sexual recidivism among juvenile offenders. Sexual Abuse: A Journal of Research and Treatment 19 (2), 107 113. Campos, P., Saguy, A., Ernsberger, P., Oliver, E. & Gaesser, G. (2006). The epidemiology of overweight and obesity: Public health crisis or moral panic? International Journal of Epidemiology, 35 (1), 55 60. Center for Sex Offender Management. (2001). Community notification and education. Retrieved from: http://www.csom.org/pubs/notedu.html Church, W.T., Wakeman, E.E., Miller, S.L., Clements, C.B. & Sun, F. (2008). Community attitudes toward sex offenders scale: The development of a psychometric assessment instrument. Research on Social Work Practice, 18, 251 259. Cohen, S. (1972). Folk Devils and Moral Panics: The Creation of the Mods and Rockers. London: MacGibbon & Kee. Cohen, S. (2004) Folk Devils and Moral Panics (3rd Edition). London: Routledge. Cole, S. (1993). Making Science: Between Nature and Society. Cambridge, MA: Harvard University Press. 340

PAGE 341

Conley, T., Hill, K., Church II, W.T., Stoeckel, E. & Allen, H. (2011). Assessing probation and community corrections workers' attitudes toward sex offenders using the community attitudes toward sex offenders (CATSO) scale in a rural state. Sexual Addiction and Compulsivity, 18, 75 85. Davis, N.J. & Stasz, C. (1990). Social Control of Deviance: A Critical Perspective. New York: McGraw Hill. Department of Justice, Office of Justice Programs (2011). Justice Department Announces Four More Jurisdictions Implement Sex Offender Registration and Notification Act. Retrieved from : http://www.ojp.usdoj.gov/newsroom/pressreleases/2011/SMART11102.htm Devine, P.G. (1989). Stereotypes and prejudice: Their automatic and controlled components. Journal of Personality and Social Psychology, 56 (1), 5 18. Durkheim, E. (1884/2011). Durkheim's Philosophy Lectures: Notes from the LycÂŽe de Sens Course, 1883 1884 From Gross, N., Jones, R.A. and Joas, H. Cambridge: Cambridge University Press. Farrington, D. (2013). Florida lawmakers to revisit sex offender laws. Retrieved from: http:/ /www.naplesnews.com/news/2013/sep/24/florida lawmakers revisit sex offender laws/ Federoff, P., & Moran, B. (1997). Myths and misconceptions about sex offenders. Canadian Journal of Human Sexuality, 6 263 277. Ferguson, K., & Ireland, C. A. (2006). Att itudes towards sex offenders and the influence of offence type: A comparison of staff working in a forensic setting and students. The British Journal of Forensic Practice, 8, 10 19. Ferraro, K.F. (1995). Fear of Crime: Interpreting Victimization Risks. Albany, NY: SUNY Press. Ferraro, K .F. (1996). Women's fear of victimization: Shadow of sexual assault? Social Forces, 75 (2), 667 690. Ferraro, K.F. & LaGrange, R. (1987). The measurement of fear of crime. Sociological Inquiry 57 (1), 70 101 Florida Department of Education. (2005). Technical Assistance Paper Jessica Lunsford Act. Retrieved from: http://info.fldoe.org/docushare/dsweb/Get/Document 3151/k12_05 107a.pdf 341

PAGE 342

Florida Department of Law Enforcement. (2012 a ). The Florida Department of Law Enforcement, Florida Sexual Offenders and Predators Website. Retrieved from: http://offender.fdle.state.fl.us/offender/homepage.do Florida Department of Law Enforcement. (2012b). Personal correspondence with the Sexual Predator Unit. Received October 24, 2012. Finkelhor, D. & Jones L.M. (2004). Explanations for the decline in child sexual abuse cases. Washington, D.C.: Office of Juvenile Justice and Delinquency Prevention. Finkelhor, D. & Jones, L.M. (2012). Have sexual abuse and physical a buse declined since the 1990s? Durham, NH: Crimes Against Children Research Center. Finkelhor, D. & Redfield, D. (1984). How the public defines sexual abuse. In David Finkelhor (ed.), Child Sexual Abuse: New Theory and Research New York: Free Press. Fisher, B.S. & Sloan, J.J. (2003). Unraveling the fear of victimization among college women: Is the "shadow of sexual assault hypothesis" supported? Justice Quarterly, 20, 633 659. Gainesville Police Department. (2012). Sexual Offenders. Retrieved from: https://www.gainesvillepd.org/index.php/crime prevention/sexual offenders Gainesville Sun. (2012). Online link to the Alachua County Sheriff's Office Publication, Alachua County's Registered Sexual Offenders and Predators Retrieved from: http://www.g ainesville.com/assets/pdf/2012_Alachua_County_Sexual_Offenders. pdf Garland, D. (2008). On the concept of moral panic. Crime, Media, Culture, 4 (1), 9 30. Goldberg, P. (1980). The federal government's response response to illicit drugs, 1969 1978. The Drug Abuse Council The Facts About "Drug Abuse." New York: Free Press, pp. 20 62. Gonnerman, J. (2007). The house where they live. Retrieved from: http://nymag.com/news/features/42368/ Goode, E. & Ben Yehuda, N. (1994). Moral Panics: The Social Construction of Deviance. Cambridge, MA; Blackwell Publishers, Inc. Gorham, B.W. (2006). News media's relationship with stereotyping: The linguistic intergroup bias in response to crime news. Journal of C ommunication, 56, 289 308. 342

PAGE 343

Greensberg, B.S., Mastro, D., & Brand, J.E. (2002). Minorities and the mass media: Television into the 21 st Century. In J. Bryant and D. Zillman (Eds.), Media Effects: Advances in Theory and Research (2 nd ed., pp. 333 352). Ma hwah, NJ: Erlbaum. Gr iffin, M. & West D. (2006). The lowest of the low: Addressing the disparity between community view, public policy, and treatment effectiveness for sex offenders. Law & Psychology Review, 30 143 169. Gusfield, J.R. (1955). Social structure and moral reform: a study of the Women's Christian Temperance Union. American Journal of Sociology, Vol. 61, pp. 221 232. Hall, S., Critcher, C., Jefferson, T., Clarke, J. & Roberts, B. (1978). Policing the Crisis: Mugging, the State, and Law and Order. London: Macmillian. Harris, P.M. (2006). What community supervision officers need to know about actuarial risk assessment and clinical judgment. Federal Probation Vol. 70, 2, pp. 8 14. Hilinski, C.M. (2009). Fea r of crime among college students: A test of the shadow of sexual assault hypothesis. American Journal of Criminal Justice, 34, 84 102. Hogue, T. E. (1993). Attitudes towards prisoners and sexual offenders. Issues in Criminal & Legal Psychology, 19, 23 32. Hollinger, R. & Lanza Kaduce, L. (1989). The process of criminalization: The case of computer crime laws. Criminology, 26 (1), 101 126. Ipeirotis, P. (2013). "Demographics of Mechanical Turk." Retrieved from: http://archive.nyu.edu/bitstream/2451/29585/2/CeDER 10 01.pdf Jacob Wetterling Act of 1994, Public Law 103 322. (1994). Retrieved from: http://www.csosa.gov/about/policies/docs/4006_sex_offender_registration_10310 3.pdf Jacob Wetterling Resource Center (2010). Who We Are. Retr ieved from: http://www.jwrc.org/ Jenkins, P. (1998). Moral Panic: Changing Concepts of the Child Molester in Mode rn America. Yale University Press. Jones, B.J., Gallagher, B.J., & McFalls, J.A. (1989). Toward a unified model for social problems theory. Journal for the Theory of Social Behavior, Vol. 19, pp. 337 356. Jones, R .A (1999). The Development of Durkheim's Social Realism. Cambridge: Cambridge University Press. 343

PAGE 344

Klein, J.L., Zambrana, K. & Rukus, J. (2012). Do societal reactions lead to increased experiences of shame and strain for registered female sex offenders? Justice Policy Journal, forthcom ing. Kobler, J. (1973). Ardent Spirits: The Rise and Fall of Prohibition. New York: G.P. Putnam. Lam, A., Mitchell, J. & Seto, M.C. (2010). Lay perceptions of child pornography offenders. Canadian Journal of Criminology and Criminal Justice, Vol. 52, 2, pp. 173 201. Lane, J., Gover, A.R. & Dahod, S. (2009). Fear of violent crime among men and women on campus: The impact of perceived risk and fear of sexual assault. Violence and Victims, 24 (2), 172 192. Lehigh University (2006). Sexual Assault Awaren ess Month events aim to educate students. Retrieved from : http://www3.lehigh.edu/News/V2news_story.asp?iNewsID=1737 Levenson, J. & Cotter, L (2005). The effect of Megan's Law on sex offender reintegration. Journal of Contemporary Criminal Justice, 21 (1), 49 66. Levenson, J., D'Amora, D. & Hern, A. (2007). Megan's law and its impact on community re entry for sex offenders. Behavioral Sciences and the Law 25 (4), 587 602. Levenson, J., Zgoba, K., & Tewksbury, R. (2007). Sex offender residence restric tions: Sensible crime policy or flawed logic? Federal Probation 71 ( 3 ), 2 9. Lieb, R. & Nunlist, C. (2008). Community notification as viewed by Washington's citizens: A ten year follow up. (Report no. 08 03 1101). Olympia, WA: Washington State Institut e for Public Policy. Liska, A.E., Sanchirico, A. & Reed, M.D. (1988). Fear of crime and constrained behavior: Specifying and estimating a reciprocal effects models. Social Forces, 66, 827 837. Logan, W. A. (2003). Sex offender registration and community notification: Emerging legal and research issues. Annals of the New York Academy of Sciences 989 337 351. Logan, W. (2007). Sex offender registration and community notification: Past, present, and future. New England Journal on Criminal and Civil Confinement, 34 3 16. 344

PAGE 345

Lonsway, K.A. & Archambault, J. (2012). The "justice gap" for sexual assault cases: Futu re directions for research and reform. Violence Against Women, 18 145 168. Madden, S. (2008). The labeling of sex offenders. Lanham, MD: University Press of America. Maguire, K. & Pastore, A.L. (2003). Sourcebook of criminal justice statistics 2003. W ashington, D.C.: Department of Justice, Bureau of Justice Statistics. Mancini, C. (2013). Sex Crime, Offenders & Society : A Critical Look at Sexual Offending and Policy. Durham, North Carolina: Carolina Academic Press. Maynard, C. & Wiederman, M. (1997). Undergraduate students' perceptions of child sexual abuse: Effects of age, sex, and gender role attitudes. Child Abuse and Neglect, 21 833 844. Mea rs, D. P., Hay, C., Gertz, M., & Mancini, C. (2007). Public opinion and the found ation of the juvenile court. Criminology 45 223 258. Mears, D.P., Mancini, C., Gertz, M. & Bratton, J. (2008). Sex crimes, children and pornography: Public views and public policy. Crime and Delinquency, Vol. 54, pp. 532 559. Megan Kanka Foundation (2001). Our Mission. Retrieved on November 8, 2010: www.megannicolekankafoundation.org/ Megan's Law of 1996, Public Law 104 145. (1996). Retrieved from: http://www.gpo.gov/fdsys/pkg/BILLS 104hr2137enr/pdf/BILLS 104hr2137enr.pdf Melvin, K. B., Gramling, L. K., & Gardner, W. M. (1985). A scale to measure attitudes toward prisoners. Criminal Justice and Behavior, 12, 241 253. Michigan State Police. (2006). Sex offender registry: Summary of legislation. Retrieved from: http://www .michigan.gov/documents/SOR_Legislation_Summary_122883_7.pdf Minnesota Department of Corrections (2003). Level 3 sex offenders residential placement issues Minneapolis, MN: Minnesota Department of Corrections. Montana's Sexual or Violent Offender Regis tration Act. (1995). Retrieved from: http://data.opi.mt.gov/bills/mca_toc/46_23_5.htm Monteith, M. (1993). Self regulation of prejudiced responses: Implications for progress in prejudice reduction efforts. Journal of Personality and Social Psychology, 65 469 485. 345

PAGE 346

MSNBC News Services (2008). Police: 1981 Murder of Adam Walsh Solved. Retrieved from: http://www.msnbc.msn.com/id/28257294/ National Conference of State Legislatures (NCSL). (2008). Adam Walsh child protection and safety act of 2006. Retrieved from: http://www.ncsl.org/IssuesResearch/CivilandCriminalJustice/HR4472AdamWalsh ChildProtectionandSafetyAc/tabid/12699/Default.aspx N elson, E.D. (1994). Females who sexually abuse children: A discussion of gender stereotypes and symbolic assailants. Qualitative Sociology, 17 (1), 63 88. New Jersey State Police. (2012). Office of the Attorney General, Sex Offender Registry. Retrieved from: http://www.nj.gov/njsp/spoff/faq.html Ost, S. (2002). Children at risk: Legal and societal perceptions of the potential threat that the possession of child pornography poses to society. Journal of Law and Society, 29 (3), 436 460. Panzati, A. & Treen, D. (2013). Donald Smith indicted on charges for murder, kidnapping, and sexual battery. Retrieved from: http://members.jacksonville.com/news/crime/2013 07 02/story/donald smith indicted charges murder kidnapping and sexual battery Petrosino, A & Petrosino, C (1999). The public safety potential of Megan's Law in Massachusetts: An assessment from a sample of criminal sexual psychopaths. Crime and Delinquency 45 (1), 1240 158. Phillips, D. M. (1998). Community notification as viewed by Washington's citizens Olympia: Washington State Institute for Public Policy. Potter, R.H. & Potter, L.A. (2001). The internet, cyberporn and sexual exploitation of children: Media moral panics and urban myths for middle class parents?. Sexuality and C ulture, 5 (3), 31 48. Pratt, J. (2000). Sex crimes and the new punitiveness. Behavioral Sciences and the Law, 18 135 151. Prescott, J. & Rockoff, J. (2008). Do sex offender registration and notification laws affect criminal behavior. University of Michigan Legal Working Paper Series. Retrieved from: http://law.bepress.com.lp.hscl.ufl.edu/umichlwps/olin/art85/ Proctor, J.L., Badzinsk i, D.M., & Johnson, M. (2002). The impact of media on knowledge and perceptions of Megan's Law. Criminal Justice Policy Review, 13, 356 379. 346

PAGE 347

Rader, N.E., Cossman, J.S., & Porter, J.R. (2012). Fear of crime and vulnerability: Using a national sample of Americans to examine two competing paradigms. Journal of Criminal Justice, 40 (2), 134 141. Redlich, A. (2001). Community notification: Perceptions of its effectiveness in preventing child sexual abuse. Journal of Child Sexual Abuse 10, 91 116. Roberts, J. V., Stalans, L. J., Indermaur, D., & Hough, M. (2003). Penal populism and public opinion: Lessons from five countries New York: Oxford University Press. Ross, J., Zaldivar, A., Irani, L. & Tomlinson, B. (2010). "Who are the Crowdworkers?: Shifting demographics in Amazon Mechanical Turk." Retrieved from: http://www.ics.uci.edu/~jwross/pubs/SocialCode 2009 01.pdf Sample, L.L. & Bray, T.M. (2003). Are sex offenders dangerous? Criminology & Public Policy, 3 (1), 59 82. Sandler, J., Freeman, N., & Socia, K. (2008). Does a watched pot boil? A time series analysis of New York State's sex offender registration and notification law. Psychology, Public Policy, and Law 14 (4), 284 302. Sanghara, K.K. & Wilson, J.C. (2006). Stereotypes and attitudes about child sexual abusers: A comparison of experienced and inexperienced professions in sex offender treatment. Legal and Criminological Psychology, 11 (2), 229 244. Seto, M. & Ekem A.W. (2005). The crimina l histories and later offending of child pornography offenders. Sexual Abuse: A Journal of Research and Treatment, 17 201 211. Schram, D., & Milloy C. (1995). Community notification: A Study of offender characteristics and recidivism. Urban Policy Research: Seattle, WA. Shiavone, S.K. & Jeglic, E.L. (2009). Public perception of sex offender social policies and the impact on sex offenders. International Journal of Offender Therapy and Comparative Criminology, 53 (6), 679 695. Silman, J. (2012). AS O: 6 arrested in Web sting, there to meet underage girls for sex. Gainesville Sun, July 23, 2012. Retrieved from: http://www.gainesville.com/article/20120723/ARTICLES/120729883 Sims, B., & Reynolds, M. (2007). Sex offender registration, notification, an d civil commitment statutes: Due process vs. community safety. In C. Hemmens (Ed.), Current Legal Issues in Criminal Justice (pp. 195 204). Los Angeles: Roxbury. Sinclair, A. (1962). Prohibition: The Era of Excess. London: Faber & Faber. 347

PAGE 348

Skolnick, J.R. (2007). Racial profiling Then and now. Criminology, 6 (1), 65 70. Skolnic k J.R. (2010). A sketch of the policeman's working personality. In Michael White (ed.), Race, Ethnicity, and Policing: New and Essential Readings. Sutherland, E.H. (1950). The diffusion of sexual psychopath laws. The American Journal of Sociology 56 142 148. Sundt, J. L., Cullen, F. T., Applegate, B. K., & Turner, M. G. (1998). The tenacity of the rehabilitative ideal revisited: Have attitudes toward offender tr eatment changed? Criminal Justice and Behavior 25 426 442. Surette, R. (2011). Media, Crime and Criminal Justice: Images, Realities and Policies. Stamford, CT: Cengage Learning. Taylor, D.G., Taub, R.P. & Peterson, B. (1986). Crime, community organization and cause of neighborhood decline. pp. 161 177 in Metropolitan Crime Patterns. Robert Figleo, Simon Hakin and George Rengert (eds). Willow Tree Press. Tewksbury, R. (2004). Experiences and attitudes of registered female sex offenders. Federal Probation Vol. 68(3), 30 33. Tewksbury, R. (2005). Collateral consequences of sex offender registration. Journal of Contemporary Criminal Justice 21 67 81. Tewksbury, R. & Lees, M. (2007). Perceptions of punishment: How registered sex offenders view registries. Crime and Delinquency 380 407. Tewksbury, R., Mustaine, E.E., & Payne, B.K. (2012). Community corrections professionals' views about sex offender s: Is the CATSO applicable? Criminal Justice Studies, 25 (2), 145 157. Tewksbury, R. & Zgoba, K. (2010). Perceptions and coping with punishment: How registered sex offenders respond to stress, internet restrictions and collateral consequences of registration. International Journal of Offender Therapy and Comparative Criminology 54 (4), 537 551. University of Florida Police Department (2010). Jeanne Cler y Act/Crime Stats. Retrieved from: http://www.police.ufl.edu/pdf_files/2010/UFPD%20Safe%20Cam pus 9.pdf U.S. Census Bureau. (2006). Population distribution in 2005 Washington, DC: Author. Veysey, B., Zgoba, K., & Dalessandro, M. (2008) A preliminary step towards evaluating the impact of Megan's Law: A trend analysis of sexual offenses in New Jersey from 1985 2005. Justice Research and Policy, 10 (2), 1 17. 348

PAGE 349

Walker, S. (2001). Sense and Nonsense About Crime. Belmont, CA: Wadsworth. Warr, M. (1984). Fear of victimization: Why are women and the elderly more afraid? Social Science Quarterly, 65, 681 702. Waterman, C.K. & Foss Goodman, D. (1984). Child molesting: Variables relating to attribution of fault to victims, offenders, and no n participating parents. Journal of Sex Research Vol. 20, 329 349. WCJB 20. (2012). Parents: No note on predator teacher. February 6, 2012. Retrieved from: http://www.wcjb.com/local news/2012/02/parents no note predator teacher Welchans, S. (2005). Me gan's Law: Evaluations of sexual offender registries. Criminal Justice Policy Review 16 (2), 123 140. Wilcox, P., Jordan, C.E. & Pritchard, A.J. ( 2006 ). Fear of acquaintance versus stranger rape as a "master status": Towards refinement of the "shadow of sexual assault." Violence and Victims, 21, 355 370. Wilson VanVoorhis, C.R. & Morgan, B.L. (2007). Understanding power and rules of thumb for determining sample sizes. Tutorials in Quantitative Methods for Psychology, 3 (2), 43 50. Zanbergen, P. & Hart, T. (2006). Reducing housing options for convicted sex offenders: Investigating the impact of residency restriction laws using GIS. Journal of Research and Policy, 8 (2), 1 24. Zatz, M.S. (1987). Chicano youth gangs and crime: The creation of a moral panic. Contemporary Crises, Vol. 82, 5, pp. 1187 1207. Zevitz, R. & Farkas, M. (2000 ). Sex offender community notification: Managing high risk criminals or exacting further vengeance? Behavioral Sciences and the Law 18 (2/3): 375 391. 349

PAGE 350

BIOGRAPHICAL SKETCH Jennifer L. Klein graduated summa cum laude from the University of Florida in 2009 with a Bachelor of the Arts in criminology and a minor in history. She graduated in 2011 with a Master of Arts in criminology, law and society from the University of Florida. She also graduated in May 2014, from the University of Florida with a Ph.D in criminology, law and society, focusing on sexual offenders and the sex offender registry as evidenced by this dissertation. 350


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID ELBUV02LN_I5XFSK INGEST_TIME 2014-10-03T21:42:19Z PACKAGE UFE0046459_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES