|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 COMPARISON OF MMPI A CHARACTERISTICS OF ADOLESCENT SEX OFFENDERS AND JUVENILE DELINQUENTS B y AMANDA N. SCHWAIT A THESIS PRESENTED TO THE GRADUATE S CHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FUL F ILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2011
2 2011 Amanda N. Schwait
3 ACKNOWLEDGEMENTS I could not have done any of this without my amazing support system : I thank my parents, Shannon, and Leah for their unending support, love, and care. I am forever appreciative of Dr. Herkov having faith in my ability to succeed. And last, but not even close to least, m y cohort (and Esther) have been the most amazing group I could have imagined with which to travel this road. I only hope to be able to find words and actions to express my gratitude to all of those who have gifted to me their kindness, compassion, and support.
4 TABLE OF CONTENTS page ACKNOWLEDGEMENTS ................................ ................................ ............................... 3 LIST OF TABLES ................................ ................................ ................................ ............ 5 LIST OF FIGURE S ................................ ................................ ................................ .......... 6 LIST OF ABBREVIATIONS ................................ ................................ ............................. 7 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 11 2 INSTRUMENTS AND METHODS ................................ ................................ .......... 23 Pa rticipants ................................ ................................ ................................ ............. 23 Instruments ................................ ................................ ................................ ............. 24 Procedure ................................ ................................ ................................ ............... 25 3 RESULTS ................................ ................................ ................................ ............... 27 4 DISCUSSION ................................ ................................ ................................ ......... 32 LIST OF REFERENCES ................................ ................................ ............................... 38 BIOGRAPHIC AL SKETCH ................................ ................................ ............................ 41
5 LIST OF TABLES Table page 3 1 Results of ANOVA for Basic Scales ................................ ................................ ... 28 3 2 Results of ANOVA for Harris Lingoes Subscales ................................ ............... 29
6 LIST OF FIGURES Figure page 3 1 Comparison of MMPI A profiles (basic scales) ................................ ................... 30 3 2 Comparison of Harris Lingoes profiles ................................ ............................... 31
7 LIST OF ABBREVIATIONS D Depression D 1 Subjective Depression D 2 Psychomotor Retardation D 3 Physical Malfunctioning D 4 Mental Dullness D 5 Brooding Hy Hysteria Hs Hypochondriases Ma Hypomania Ma 1 Amorality Ma 2 Psychomotor Acceleration Ma 3 Imperturbability Ma 4 Ego Inflation Mf Masculinity/Femininity Pd Psychopathic Deviance Pd 1 Familial Discord Pd 2 Authority Problems Pd 3 Social Imperturbability Pd 4 Social Alienation Pa Paranoia
8 Pt Psychasthenia Sc Schizophrenia Si Social Introversion Si 1 Shyness/Self Consciousness Si 2 Social Avoidance Si 3 Alienation Self and Others
9 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science COMPARISON OF MMPI A CHARAC TERI S TICS OF ADOLESCENT SEX OFFENDERS AND JUVENILE DELINQUENTS By Amanda N. Schwait December 2011 Chair: Michael J. Herkov Major: Psychology In this study, scores on the Minnesota Multiphasic Personality Inventory for Adolescents (MMPI A) were compared between two groups of juvenile delinquents: those who have been charged with sexually based offenses and those who have not. Through analysis of personality profiles as assessed by this test, this study sought to determine what differences there were, if any, between adolescents who have been adjudicated for sexual offenses as opposed to their non sex offending peers. The final sample consisted of 542 juvenile delinquents, ages 12 18, involved in Department of Juvenile Justice in northe ast Florida. This sample was divided into sex offenders (n=79) and non sex offenders (n=463) based on current and past charges. We hypothesized that sexual offenders would have higher scores on scales indicating difficulty relating to peers (e.g. depression, social introversion) and that non sex offenders would score higher on measures indicating difficulty aligning with authority. Results of statistical analysis supported our first hypothesis, revealing two significant differences between delinquent groups o n scale 0 (Social Introversion) and scale 7 (Psychasthenia). There were also notable differences on several of the Harris Lingoes subscales, further
10 indicating that sexual offenders exhibit more shy and introverted personalities than their non sex offendin g counterparts. There were no differences directly supporting our second hypothesis. These findings indicate that there may be salient differences in peer interaction styles between these two groups of offenders. With this knowledge, assessment and treatme nt could focus more attention on the salient personality characteristics found when using broad measures of personality and address the social introversion found in sex offenders.
11 CHAPTER 1 INTRODUCTION There are few crimes in the United States that are viewed as negatively as sexual offenses. Stories of sexual offenses especially those involving children elicit strong visceral reactions from the public. Research indicates that aberrant sexual behavior, which in many cases is associated with psychopathy of personality, is thought to be established early in life, be relatively stable, and to ha ve a long duration (Lynam et al. 2009). While the prototypical sexual offender is often portrayed as an older male, sexual offenders are likely to have committed t heir first offense by the age of 14 (Freeman, Dexter Mazza, & Hoffman, 2005) and Davis and Leitenberg (1987) found that nearly half of adult sex offenders reported committing their first offense during adolescence. Thus, it is not surprising that adolescen ts account for a relatively large proportion of the sexual offenses committed in the United States. The Federal Bureau of Investigation (2009) records indicate that males under the age of 19 are responsible for between 10% and 15% of all forcible rapes, w hile 18% of all other sex crimes (e.g. molestation, non contact offenses, etc.) are perpetrated by individuals under the age of 18. In addition, individuals between the ages of 18 and 21 perpetrate nearly 30% of all forcible rapes and 30% of all other sex crimes. Moreover, research has shown that the pattern of deviant sexual behavior remains stable over time once these behaviors ar e established (e.g. Blaske, Borduin, Hennegler, & Mann, 1989, Lynam et al 2009). Therefore, within psychological and legal co mmunities, there exists a sense of urgency in becoming more keenly focused on adolescent sexual offenders. More specifically, understanding the personality dynamics and behavioral patterns within this group can lead to elucidating more effective preventati ve and intervention programs. This study seeks to compare
12 the personality characteristics of a group of adolescent sexual offenders with a group of juvenile delinquent peers who have not committed sexually based offenses. This research will help to unde rstand more about common personality traits in adolescent sex offenders and how they differ from other delinquent groups. Making concrete assumptions about adolescent sex offenders as a whole is difficult because, like many criminal populations, they are a heterogeneous group. Several studies have attempted to classify adolescent sexual offenders into groups based on personality assessment or type of sexual offense committed. Oxnam and V ess (2008), for example, found four distinct categories of sexual offe nders, including passive aggressive, who exhibit immature and dependent personalities, and inadequate, who are experiencing a great deal of internalized emotional distress. Certain personality factors have repeatedly been established in the literature. Re search indicates that juvenile sex offenders tend to exhibit poor social interaction with peers and general social isolation (e.g. Gudjonsson & Sigurdsson, 2000, Freeman et al. 2005). They often suffer from difficulty in academic settings, either due to le arning disabilities or emotional difficulty (Freeman et al 2005). Rich (2003) indicates that the majority of sexual offending adolescents have low to average IQ levels. Even between the categories of sex offenders found by Oxnam and Vess (2008) there are commonalities of poor peer relationships and difficulties with social rules. Rich (2003) suggests that while this population is undoubtedly heterogeneous, there are underlying commonalities of adolescent sexual offenders that are typically found. These fa ctors include a history of personal victimization, mental health issues, limited empathy, underdeveloped social skills, regressed moral development, and witness to family
13 dysfunction (Rich, 2003 ) It is unclear, however, if these qualities can also typical ly be found in the nonsexual offending population of juvenile delinquents or if they are specific to those who offend sexually Research ha s identified the role childhood physical and sexual abuse plays in the development of these deviant behaviors (e.g. Jonson Reid & Way, 2001 Rich, 2003 ), with about 28% of sexual offenders reporting sexual victimization in their childhood (Langevin, Handy, & Wright, 1989). Oxnam and Vess (2008) reported that between 37% and 72% of their adolescent sexual offending popul ation were physically abused by a parent or guardian during childhood. However, the majority of childhood abuse victim s do not go on to commit sexually based offenses and many sexual offending adults do not have a history of sexual or physical victimizatio n. Thus, gaining insight into other factors (particularly those that are based on personality traits) tha t influence sexual offending will allow for a more complete understanding of adolescent sex offenders and their motivations. Additionally, more efficac ious treatment plans for those who have committed these crimes can be developed if clinicians know how sex offenders compare to their non sex offending peers and on what areas are most important to focus. There is a considerable body of research on the per sonality characteristics of adult sexual offenders. More specifically, researchers have examined differences and similarities in measures of personality between groups of adult sex offenders and those that commit other criminal acts For example, Gudjonsson and Sigurdsson (2000) found that adult offenders of non sex based crimes demonstrated more extraversion and exhibited more aggression than their sexual offending counterparts. Results of a
14 study by Craig, Browne, Beech, and String er (2006) indicated that their sample of sex offenders showed fewer impulsive behaviors and less aggressive ideation than violent offenders whose crimes were not sex based. While these results are important in profiling and treating adult offenders, more r esearch with adol escent populations is valuable in order to see if these characteristics are also observed in adolescence. It would be in poor practice to assume that the research findings found in adult sexual offenders are able to generalize to adolescen ts. While the behaviors within the sexual crimes of adults and adolescents may be similar, their motivations and reasoning behind offending may be very different (Rich, 2003). Therefore, only recently more studies have been dedicated to looking at devia nt adolescent populations in order to ascertain if the characteristics found in adults are visible in adolescent offenders. The MMPI, created in 1943 by Hathaway and McKinley, was originally designed to help psychologists appropriately diagnosis patients (Graham, 2006) Since its inception and revision in 1989, it has been readily used as an assessment tool to help augment knowledge of personality characteristics and symptoms of emotional distress (Graham, 2006). It has been used in a variety of settings i ncluding the legal system, mental health hospitals, and v ocational rehabilitation sites. Within the legal system, it has been used in many contexts, including but not limited to classification of sexual predators, evaluations of competency, and aiding in d etermination of sentencin Pennuto & Archer, 2 008). In research, the MMPI has been a useful and efficient tool to assess personality deviance and make comparisons between groups. For example, early research by Hathaway & Monachesi (1957) suggeste d that scales 4 (Psychopathic Deviance), 8 (Schizophrenia), and 9 (Hypomania), were "excitatory" scales associated
15 with a high risk of delinquency. Conversely, scales 0 (Social Introversion), 2 (Depression), and 5 (Masculinity Femininity), were "suppressor scales associated with a lower risk of delinquency. Elevations on Scales 1 (Hypochondriasis), 3 (Hysteria), 6 (Paranoia), and 7 (Psychasthenia), were found to not be associated with delinquency. Although some studies have shown some evidence against the ideas posed by Hathaway & Monachesi (1957) (e.g. Archer, Bolinskey Morton, & Farris, 2003), these early studies exemplify the rich history the MMPI has had in research. Specifically, the ability to use the MMPI to distinguish differences and similarities between groups has made it vastly popular in research settings. Research has shown that the adolescent version of the MMPI (MMPI A) is the most widely used self report measure in assessing the adolescent population (Baum, Archer, Forbey & Handel, 2009). Historically, the original version of MMPI was used with adolescent populations. However, this proved to be problematic, as the normative samples for the MMPI did not contain an adolescent sample. Over the years, adolescent norms for use with the adult MMP I were developed (e.g., Colligan & Offord, 1987; Gottesman et al 1987). Despite the addition of this normative sample there were still significant problems with using this adult version of the MMPI in this population. First, some c l inicians mistakenly used the adult norms with adolescents resulting in elevations on certain scales and inaccurate description of adolescent characteristics. Second, many clinicians felt that the MMPI had inappropriate content for adolescents and was not relevant in descr ibing the adolescent experience (Archer & Krishnamurthy, 2002) Finally, these adolescent norms did not apply to all scales, resulting in limited utility of the MMPI. All of these factors led to the development of an
16 adolescent version of the MMPI. Seve ral changes were enacted in order to make this test better suited for adolescent use. For example, the complete version of the MMPI A consists of only 470 items, nearly 100 less items than the MMPI and the primary test validity and clinical scales can be s cored using only the first 350 questions While many of the questions on the test are the same as those of the MMPI 2, other q uestions were developed to pertain more to everyday experiences of adolescents (e.g. peer relationships, drug use) and some items with difficult or outdated language were dropped (Archer & Krishnamurthy, 2002). However, despite the differences, care was taken to ensure continuity between the MMPI and the adolescent version maintaining the psychometric properties that have made the M MPI such a popular assessment tool Due to the development of a specific adolescent version of the MMPI, it is posited that more productive research in adolescent personality patterns can now be conducted with more valid results (Archer, 1997). Within the wider group of juvenile delinquents, several patterns of personality profiles on the MMPI A have been described the most consistent being elevations in scale 4 (Psychopathic Deviance) (e. g. Archer et al. 2008). Elevation s on Psychopathic Deviance indicate difficulty aligning with the values of society and exhibition of deviation from typical socially acceptable behaviors While moderate elevations on scale 4 can logically be expected across varying groups of adolescents ( due to the struggle many adolescents have with authority), these elevations are more pronounced in those adolescents who have been adjudicated of a legal offense. Morton, Farris, and Benowitz (2002) found strong evidence to suggest that elevations on scale s 4, Psychopathic Deviance, and 6, Paranoia, are the most
17 characteristic patterns of scoring for juvenile delinquents in their population of adolescent s involved in the juvenile legal system. Other patterns and predictors have been suggested in the literat ure, such as the research done by Parker, Morton, Lingefelt, and Johnson (2002), which found that of the basic scales of the MMPI A, only scale 3, Hysteria, predicted future violent offenses in their population. Despite the wide use of the MMPI with adoles cent populations, there is a dearth of research looking at adolescent sex offenders and how they score on this measure. In comparing groups o f juvenile delinquents to non delinquent peers, using a broad definition of delinquency may be appropriate. Howeve r, it is logical to suppose that crimes vary widely in their levels of severity and that those who commit the most serious offenses (e.g. assault on a person) may differ from those who commit non violent and less severe offenses (e.g. drug possession). A s tudy by Glaser, Calhoun, & Petrocelli (2002) exemplifies this idea of offense classification within the larger group by grouping together adolescents based on type and severity of crime. This study specifically looked at those adolescents who perpetrated c rimes against people, property offenses, and drug/alcohol offenses and sought to determine if differences were apparent between their MMPI A profiles ( Glaser et al., 2002) Descriptive discriminant analysis of these groups and their relative MMPI A profile s indicated that the groups had small, but significant differences in their characteristics. Those delinquents who were more actively concerned with bodily functions and somatic concerns were less likely to have been adjudicated of drug/alcohol charges and were more likely be involved in property offenses (Glaser et al 2002) This potentially could be because, despite their delinquent nature, they are unlikely to involve themselves in
18 behaviors that would deliver uncomfortable bodily consequences (as might occur with drug or alcohol use). Such research studies as this demonstrate the importance of looking at within group differences among juvenile delinquents. In a study on MMPI A scor e differences between three male adolescent groups, a standard clinical control group, a delinquent group, and a dual diagnosis group (e.g. psychiatric and substance abuse diagnoses), elevations in scale 4 (Psychopathic Deviance) T scores were found for all groups, which was consistent with prior research ( Archer et al., 2003) However, their MANOVA results also showed significant differences in scoring patterns between juvenile delinquents and those adolescents in the standard clinical group (e.g. those receiving psychiatric treatment in an inpatient facility). Results from their MANOVA found significant differences in mean scores for Hypochondriasis, Psychopathic Deviance, Masculinity/Femininity, Paranoia, and Social Introversion. While these score diff erences were statistically significant, the magnitudes of many of these differences were not particularly robust (e.g. differences of 5 T score points or less). Through analysis of MMPI A scores, members of the delinquent group were found to demonstrate gr eater social immaturity and appeared to resist admitting psychological distress. Members of the groups involved in psychiatric treatment exhibited more suspicion and concern about the motives of others. However, the relatively small differences in T score s found in this study could be attributed to the use of three inpatient facility groups, as opposed to groups from who are not involved in treatment. These results indicate that the MMPI A has utility in classifying clinical adolescent populations, but tha t the differences between groups may not be as large as they originally expected (Archer et al. 2003)
19 There is a history of research using the MMPI as a personality measure within the specific criminal population of adult sexual offenders. Early research by Armentrout & Hauer (1978) suggested that there were specific codetypes that were found in groups of sexual offenders despite against whom their crime was committed (e.g. children, adults), finding elevations in scale 4, Psychopathic Deviance, and scal e 8, Schizophrenia. This research did note that profiles of those who offended against adults indicated more hostility and feelings of resentfulness than those who offended against children (Armentrout & Hauer, 1978) Hall, Shepherd, & Mudrak (1992) found no significant differences in profiles generated by the MMPI when comparing sexual and non sexual offenders who were both involved in crimes against children. This research further suggested that within group differences in personality profiles of sex off enders may be more pronounced than comparison to non sexual delinquents (Hall, She pherd, & Mu drak, 1992). Davis and Archer (2010) review of personality patterns in this population l arge effect sizes in distinguishing between sex offending and control populations. However, they note that this could be due to many studies using comparison of sex offenders to non delinquent populations and the possibility that these differences represen t the deviant personality characteristics of delinquents, not characteristics specific to sex offenders. There is a paucity of research using the MMPI with adolescent sex offenders. One of the most relevant studies to the current analysis was conducted by Losada Paisey (1998), who used the MMPI A to differentiate between adolescent male sexual offenders and juvenile delinquents adjudicated of crimes other than sexual offenses By
20 looking at the most frequent scale elevations in their sample (scales that exceeded T scores of 65), Losada Paisey (1998) examined whether there were differences in the clinical elevations of certain scales. Results indicated that in the non sex offender group, scale 9 (Hypomania) and scale 7 (Psychasthenia) were the most common ly elevated groups. In the sexual offender group, scale 9 (Hypomania) and scale 8 (Schizophrenia) were the most common elevations. This study concluded that while there are some similarities between these groups, significant differences occur in areas me asuring hysterical disorders, anxiety disorders, antisocial personality characteristics, and cognitive organization (Losada Paisey, 1998). However, this sample size was small with only 51 total participants and only 21 adolescent sex offenders. Thus, this study has limited power and results should be viewed with caution. A study by Freeman et al (2005) found that their sample of non sexual offending delinquents had clinically significant scores on scale 4, Psychopathic deviance, while the juvenile sexual offenders in their sample did not. However the mean scores on this scale for groups were not statistically significant in their difference. Rich (2003) suggests that juvenile sexual offenders do not as of ten present with the behaviors associated with conduct disorders (e.g. problems with authority, confrontational behavior) as their non sex o ffending delinquent peers. In research by Herkov, Gynther, Thomas, & Myers (1996), significant differenc es were found within a population that included multiple groups of sexual offend ers. Specifically, this study found differences between groups of sexual offending adolescents, such that there were significant differences in two point codetypes between the rape/sodomy groups and the group that had committed other
21 types of sexual offenses (Herkov et al.1996) Their results suggested considerable variability within the group of adolescent sex offenders. With regard to personality differences between juvenile delinquents and sex offenders, studies have found response pattern differences in self report measures that could provide insight into demonstrated within group differences, as measur ed by personality assessments, in juvenile delinquents. One such study found evidence that sexual offenders, as opposed to other ty pes of offenders, had more severely disturbed emotional functioning (Blaske et al. 1989) Although they did not use the MMPI A for their analysis, the self reported characteristics from this study are nonetheless important in the discussion of personality differences. Using self report measures such as the Family Adaptability and Cohesion Evaluation Scales II and the Unrevealed Differences Questionnaire Revised results showed that adolescent offenders (non sexual) had several significant patterns of distur bances within their family and peer relationships (Blaske et al. 1989) These individuals were found to be disengaged in familial relations, resistant in their adaptation to novel situations, and experienced more disturbances with family and peers, inclu ding difficulty communicating (Blaske et al. 1989) In contrast, sex ual offending participants reported feeling more disconnected from their peers and experiencing more anxiety than their delinquent counterparts who were not sexual offenders. Their disturb ances tended to revolve around emotional functioning and relating to their peers, perhaps indicating lower social competency. These results indicate that the motivations behind deviant behavior may be related to the individual experience of emotions and ab ility to relate to peers and family. These experiences may partly be responsible for the type of offense the adolescent commits.
22 As summarized previously, more direct attention needs to be given to personality dimensions, especially those measured by the MMPI A, apparent in juvenile sex offenders. As noted in much of the research, there may never be a specific profile that can be assumed for all adolescent sexual offenders. However, gaining insight into underlying commonalities and possible motivations for offending using widely available personality measures will allow for more cost and time effective treatment. The re is some evidence (e.g. Blaske, 1989, Archer et al., 2003) that adolescent sex offenders may have difficult in their interactions with peers, especially in emotional regulation in comparison to a non deviant population. However, it is unclear whether this remains true when comparing the sexual offenders to non sex offending juvenile delinquents The present study is an attempt to further unders tand the differences between adolescent sex offender groups and their non sex offending counterparts. Our hypotheses for this study stem from the previous research done on populations of deviant adolescents, as well as generalizations from research conduct ed on the larger population of adult sexual offenders. In line with previously outlined research, the current study seeks to examine two main hypotheses: 1) Adolescent sex offenders will score higher than their non sex offending peers on those scales that indicate difficulty relating to peers and maladaptive social behaviors such as measures of social introversion (scale 0 Social Introversion) and anxiety (scale 7 Psychasthenia) and 2) Juvenile delinquents who have not been involved in a sexual offense w ill score higher on scales that indicate difficulty with authority and frequent conflict with family and peers including measures of impulsivity (scale 9 Hypomania) and conflict with authority (scale 4 Psychopathic Deviance)
23 CHAPTER 2 INSTRUMENTS AND METHODS Participants Data from 542 adolescent (individuals aged 12 18) participants were analyzed in this study. In accordance with Florida Statute 985.18 (Fla. Stat. 2010) prior to being admitted to a juvenile delinquent treatment facility, all i ndividuals under the age of 18 who have committed a crime may be given a psychological evaluation The information used for this archival analysis was one part of this larger evaluation. Participants in this study adolescents who where evaluated in accor d ance with the above statute from the Department of Juvenile Justice District 4 catchment area (northeast Florida). All of the youth had been adjudicated guilty of committing their particular offense and were undergoing eval uation prior to placement. The advisor for this research, Dr. Herkov, served as an evaluator in some of these cases. Because this research represents archival research, no informed consent was obtained from any of the adolescents, parents or guardians of the reviewed files. However, Institutional Review Board (IRB) approval for the project was obtained. The files were reviewed and an electronic database was developed that included demographic, offense, symptom and test data. These data were did not include any identifying information and there is no way to link the data set back to the original file. Subject files were divided into two distinct groups: the first group is comprised of those adolescents who have previously been arrested for sexually related offenses (n=79), while the se cond group is comprised of adolescents who have been arrested for nonsexual offenses (n=463). The non sex offenders were adjudicated of a variety of crimes including those involving drugs, weapons, theft, property damage, animal
24 cruelty, and non sexual att ack on a person. None of the participants in the non sex offenders group had been previously adjudicated for any sex crime. We conducted an ANOVA to assess whether a significant difference in age existed between the two groups. We found a small, but signif icant, difference in age F (1, 540) = 41.14, p < .001. The age range for the sex offenders was 12 17, M = 14.71, SD =1.49. The range for the non sex offenders was 12 18, M = 15.76, SD = 1.32. All of the participants in the sample are male. As this analysis w as done using archival information, information on race and/or ethnicity was not consistently reported and could not be analyzed appropriately. Instruments For the purposes of identifying personality patterns, we used the adolescent adaptation of the Min nesota Multiphasic Personality Inventory (MMPI A) The normative sample for the MMPI A includes 1,620 adolescents: 805 males and 815 females between 14 and 18 years old (Archer & Krishnamurthy, 2002). The completion of the full form of the MMPI A, comprise d of 470 items, determines the 10 original standard scales, as well as validity scales, content scales, Harris Lingoes subscales, supplementary scales, and the PSY 5 (i.e., five factor personality scales) The assessment has been found to have psychometric properties much like the original MMPI, with test retest correlations between .65 and .84, and a standard error of measurement between 4 and 6 T score points (Archer & Krishnamurthy, 2002). The reading level needed for the items is between fifth and seven th grade. The participants in the present study completed the firs t 350 items of the MMPI A. The entire test was not administered because many of the participants had a difficult time completing the entire test. The first 350 items can be used to determ ine the basic
25 clinical scales: Hypochondriasis Depression, Hysteria, Psychopathic Deviance, Masculinity/Femininity, Paranoia, Psychasthenia, Schizophrenia, Hypomania, and Social Introversion. It also can be used to determine the Harris Lingoes Subscales, which are further breakdowns the following basic scales: Depression, Hysteria, Psychopathic Deviance, Paranoia, Schizophrenia, Hypomania, and Social Introversion. The Harris Lingoes subscales are used as a way to refine and improve on interpretation of the basic scales, as they were constructed by looking at item content and grouping those questions that seemed to express a specific trait (Graham, 2006). These scales provide more detailed information into what types of questions the participant is endorsing and, thus, can aid in interpretation. Only those subscales that were expected to have differences consistent with our a priori hypotheses were compared between groups. Based on our hypotheses, we compared the following Harris Lingoes sub scales: Psychopathic Deviance, Hypomania, and Social Introversion. Procedure The current study uses archival data to look at differences in MMPI A responses between two groups of juvenile delinquents. The data was collected from adolescents involved in De partment of Juvenile Justice treatment programs in northern Florida in the years 2003 and 2004. As the archival data collected for the present study only consisted of the first 350 items of the MMPI A a more limited number of scales can be analyzed with s ufficient pow er. S cale comparisons and item analysis comparisons will only take into account those scales that are completely determined by the initial 350 items on the test. For purposes of validity, we imposed several exclusion criteria. The MMPI A cont ains validity scales, designed to detect random responding, exaggeration of symptoms, or minimization of distress. In clinical settings, if these scores are
26 significantly elevated (T score greater than 80) the profiles are considered invalid and results wi ll not be analyzed. All adolescents were administered a reading test as part of their clinical evaluation (Wide Range Achievement Test Third Addition). Those who did not demonstrate sufficient reading ability were administered the test via audiotape. Exc lusionary criteria involved removal of any MMPI A protocol that had any of the following validity indices with scores greater than 80: L (lie), K ( correction/ defensiveness), F (infrequency), and VRIN (variable responding), yielding a final sample of 542 p articipants.
27 CHAPTER 3 RESULTS For our analysis, we performed a one way ANOVA on the group of sex offenders and non sex offenders for the basic scales and Harris Lingoes subscales of the MMPI A. The data met all ANOVA assumptions for normality and equali ty of variances. In order to account for the large number of comparisons performed, Bonferroni corrections were applied to control for alpha inflation. Effect sizes for the differences in mean d Of th e 10 basic scales of the MMPI A, there were significant differences in mean T score on scale 0 (Social Introversion) and scale 7 (Psychasthenia). There were no other significant d ifferences between groups (T able 3 1). Sex offender participants scored highe r on Psychasthenia ( M = 50.11, SD = 11.9), than non sex offender participants ( M = 47.02, SD = 10.52), F (1, 540) = 5.61, p = .018, d = .27. Sex offender participants also scored higher on the measure of social introversion, ( M = 50.95, SD = 9.48) than non sex offender participants ( M = 47.48, SD = 9.53), F (1, 540) = 8.93, p = .003, d = .37. There were significant differences in several of the Harris Lingoes subscales, which further breaks down each basic scale based on groups of items pertaining to the same area of functioning (T able 3 2). Significant differences in means with non sex offender participants scoring significantly higher than sex offender participants occurred for Pd 1 F (1,540)= 7.27, d =.33 Pd 3 F (1,540)= 10.149, d =.37, Ma 1 F (1,540)=3.95 d =.25, and Ma 3 F (1,540)= 7.74, d =.34 (all p values less than .05) Sex offender participants scored significantly higher on Si 1 F (1,540)= .13.30, d =.42, p <.0001.
28 Table 3 1 Results of ANOVA for Basic Scales Non Sex Offenders Sex Offenders Scale M SD M SD F d Hy 50.18 10.27 50.05 9.59 0.01 0.01 D 54.73 9.7 56.85 9.73 3.23 0.21 Hs 51.78 9.54 51.42 9.44 0.09 0.04 Pd 58.15 10.13 57.28 10.86 0.48 0.08 Mf 43.00 8.60 43.70 8.79 0.45 0.08 Pa 51.21 10.51 52.05 10.40 0.43 0.08 Pt 47.02 10.52 50.11 11.90 5.61 0.27 Sc 48.66 10.64 50.56 11.24 2.10 0.17 Ma 52.24 10.88 50.85 10.82 1.10 0.13 Si 47.48 9.53 50.95 9.48 ** 8.93 0.37 Note: p < .05, **p < .001 Abbreviations: Hy = Hysteria, D = Depression, Hs = Hypochondriases, Pd= Psychopathic Deviance, Mf= Masculinity/Femininity, Pa= Paranoia, Pt= Psychasthenia, Sc= Schizophrenia, Ma= Hypomania, Si= Social Introversion
29 Table 3 2 Results of ANOVA for Harris Lingoes Subscales Non Sex Offender s n= 79 Sex Offenders Scale M SD M SD F d Pd 1 51.97 9.89 48.72 10.03 *6.65 0.33 Pd 2 61.95 8.41 61.06 8.21 0.97 0.11 Pd 3 55.57 9.69 51.75 10.78 *8.80 0.37 Pd 4 52.98 10.25 54.09 10.27 1.53 0.11 Pd 5 53.05 10.3 0 53.18 10.29 0.07 0.01 Ma 1 53.65 9.83 51.29 9.22 *4.29 0.25 Ma 2 45.43 9.36 45.39 10.69 0.15 0.00 Ma 3 57.67 9.98 54.29 9.93 *6.23 0.34 Ma 4 49.25 9.72 50.06 10.72 0.77 0.08 Si 1 44.50 8.79 48.48 9.99 **10.75 0.42 Si 2 50.11 9.89 50.53 8.59 0.08 0.05 Si 3 49.08 10.11 50.81 10.73 3.06 0.17 Note: p< .05 **<.001
30 Figure 3 1 Comparison of MMPI A p rofiles (basic s cales ) 35 40 45 50 55 60 65 Hy D Hs Pd Mf Pa Pt Sc Ma Si Mean Score Basic Scales Non-Sex Offenders Sex Offenders
31 Figure 3 2. Comparison of Harris Lingoes profiles 40 45 50 55 60 65 Pd1 Pd2 Pd3 Pd4 Pd5 Pa1 Pa2 Pa3 Ma1 Ma2 Ma3 Ma4 Si1 Si2 Si3 NonSex Offenders Sex Offenders
32 CHAPTER 4 DISCUSSION The current study was conducted in order to examine within group differences and similarities in MMPI A profiles in a population of delinquent adolescents. More specifically, the analysis sought to look at differences between groups of adolescent non sex o ffenders and adolescent sex offenders by examining their scores on the MMPI A. Of the basic scales of the MMPI A, differences were found in two scales. One of these significant differences was found in scale 0 (Social Introversion), which is an indication of an individual ability to be comfortable in a social situation, behavioral tendencies in social arenas, and interaction patterns with peers. People who have high scores on this scale tend to be reserved or shy (Archer & Krishnamurth, 2002). They are likely to be submissive and compliant in social relationships and may have difficulty forming alliances (Archer & Krishnamurth, 2002). The sex offending group scored significantly higher on this scale indicating that they felt more introverted and less com fortable with peers than the group of non sex offending adolescents. The other significant difference on the basic scales was found in scale 7, Psychasthenia. Sex offenders scored higher on this scale indicating that they tend to have more anxiety and in decisiveness. High scores on this scale can be related to obsessive and ruminative thoughts, as well as tension and apprehension (Archer & Krishnamurth, 2002). Sex offender adolescents in our study scored higher on this scale, indicating that they may expe rience more anxious symptoms and obsessive thinking that their non sex offending counterparts. Furthermore, the significant differences found in the Harris Lingoes subscales provide more detailed information into the type of personality difficulties foun d in these
33 groups. There were several significant differences, but sex offender adolescents scored higher only on Si 1 Shyness/Self Consciousness, which is comprised of items that look at feelings of self consciousness and worry in social interactions. Pe ople with high scores on this subscale tend to lack self confidence and feel anxious or uncomfortable in social situations (Archer & Krishnamurth, 2002) further indicating their relative discomfort in social situations in comparison to their non sex offen ding peers. The significant differen ces in Social Introversion and the mentioned subscales suggest that the adolescent sex offenders in our population may have a difficult time relating to their peers and feeling comfortable in social arenas, which is in support of our first hypothesis. T his finding is of key importance because m any of the individuals in our sample committed their sexual crimes against younger individuals. This could be because they lack the social maturity to engage in intimate relations hips with people their own age. It could suggest that they engage in aberrant behaviors due to their discomfort with peers and feelings of insecurity. In comparison to sex offender adolescents, the non sex offender adolescents scored higher on several o f the Harris Lingoes subscales. Notably, they scored higher on two of the subscales of Psychopathic Deviance, which measures general problems with authority and rules They reported having more problems with their family life (as measured by Pd 1, Familial Discord), which indicates that they endorsed items relating to wanting to leave their home situations and seeing their families as less loving, understanding, or supportive (Archer & Krishnamurth, 2002). Endorsement of these items may also indicat e more critical and quarrelsome views of their family life (Archer & Krishnamurth, 2002). Also from the Psychopathic Deviance subscales, non sex offender
34 adolescents scored higher on Pd 3 Social Imperturbability, which is an indication that they do not fee l as much anxiety or self consciousness in social situations. Non sex offender participants also scored higher on two subscales of Hypomania a scale that indicates a desire for action, grandiosity, and a tendency to engage in excessive activity. The non sex offenders scored higher on Ma 1 Amorality a measure of antisocial beliefs and impulsivity and Ma 3, Imperturbability a measure of one s tendency to seek out excitement and be independent from the influence of others. In support of our first hypothesi s the non sex offender adolescents in this study scored higher on Ma 3 Imperturbability, indicating that they reported feeling very little social anxiety and having no problems interacting with others. This finding is also consistent with their lower scor es on social introversion as compared to their sexual offending counterparts. Consistent with prior research (e.g. Gudjonsson & Sigurdsson, 2000, Freeman, et al. 2005 ), and with our first hypothesis, sex offender participants scored higher on scales of di fficulty interacting in social situations and introversion compared to non sex offender participants, suggesting that sex offender participants exhibited greater difficulty in interacting with their peers in social situations, and were more introverted tha n their non sex offending counterparts. However, contrary to our second hypothesis, no significant differences emerged between sex offender and non sex offender participants in this study on direct measures of having problems with authority. However, non sex offender adolescents in this study scored higher on subsca les measuring energy levels, familial discord, and
35 endorsement of antisocial ideals, such that non sex offender participants seem to be endorsing some qualities indicative of problems with autho rity. The results of this study have several implications. Significant differences on several measures of social interaction suggest that adolescent sex offenders have more difficulty in social settings than non sex offending juvenile delinquents partial ly illuminating why they engage in aberrant social beha viors, such as sexual offending They report feeling self conscious and experiencing anxiety in social situations. It is important to note that not only do they score higher on measures of social intro version, but they also score lower on scales of social comfort and imperturbability. In designing and implementing treatment programs, clinicians should focus on these areas of social functioning. Clinicians have indicated that designing individualized com prehensive treatment plans will be key in developing ways to reduce recidivism in this population (Becker and Johnson, 2001). Moreover, treatment and rehabilitation efforts could demand more focus on helping adolescent sex offenders increase their self est eem or improve their social skills by reducing social anxiety. By tailoring treatment to the specific personality problems experienced by sex offenders, we can hope to increase efficacy of treatment plans and reduce recidivism. Results do indicate deficit s in social functioning may be linked to increased risk of sexual offense and that either proactively or preemptively, social skills training may be an effective treatment modality. It is important to discuss the limitations of this study in order to redu ce misinterpretation of the data and provide a platform for further research First, in this study, we only examined data collected from males. As female delinquents were not included in the sample, we are unable to generalize our results to female adolesc ents.
36 In the future, it would be beneficial to study female sex offenders, as they have been severely underrepresented in the research regarding this population and little is known about the personality differences, if any, between female and male sex offe nders Second the data analyzed in this study w as collected as the adolescents were entering a treatment program within the Florida legal system. It is possible that they were keenly aware that their answers would be used to determine the severity of the ir treatment program thus a ffecting their answering patterns Third it is important to note that though there were statistically significant differences in group means for the scales noted, none of the means of the clinical scales obtained for any group were clinically elevated (i.e. T scores greater 65). The results are not necessarily indicative of clinically significant scores (e.g. scores that would suggest pathology) for the groups of offenders but solely that their scores showed significant deviatio n in comparison to their peers Fourth, there are inherent difficulties with using an archival data set. Our use of this data set limited our ability to ask questions for clarification on answers, to fill in missing data (e.g. ethnicity), and to tailor the information collected to the needs of the current study. Finally, it is important to note that although we treated our two groups of juvenile delinquents (sex offenders and non sex offenders) as two homogeneous groups, it is imperative to remember that th ese groups are heterogeneous. In future studies, it will be important to break down the groups to look at within group differences based on type of crime. Finding and replicating the characteristic social introversion of this population using a broad meas ure of personality such as the MMPI A now gives us a starting point to use more specific instruments to measure these characteristics. Future research
37 could focus on furthering understanding of the role of social introversion and how it manifests in adoles cent sexual offender populations. Additionally, future research should focus on bringing to light the other variables that affect sexual offending status (e.g. environmental factors, other individual difference personality variables, etc.). Therefore, fut ure studies could look at both moderating and mediating effects of other variables not directly associated with individual personality characteristics, such as exposure to abusive conditions in the home, socioeconomic status, and other environmental factor s. Though the current study focused solely on responses provided on the MMPI A, it would be illogical and nave to assume that environmental factors do not play a role in the development of sexual deviant behavior in some situations, and thus future resear ch could look to examine concomitant relations of personality and environmental factor variables. Despite its limitations, this study elucidated the possible relation between social interaction style and peer relations in adolescent sex of fenders. The comp arison between these groups of delinquent youth allows for more tailored program planning within treatment facilities for juvenile delinquents and can help to build a great body of literature discussing the specific personality characteristics apparent in juvenile sex offenders.
38 LIST OF REFERENCES Archer, R. (1997) Future directions for the MMPI A: Research and clinical issues. Journal of Personality Assessment, 68(1), 95 109. Archer, R, Bolinskey, K., Morton, T ., & Farris, K (2003). MMPI A characteristics of male adolescents in juvenile justice and clinical treatment settings. Assessment 10 (4), 400 410. Archer, R & Krishnamurth, R. (2002). Essentials of MMPI A assessment. New York, NY: John Wiley & Sons. Arment rout, J., & Hauer, A. (1978) non rapist sex offenders. Journal of Clinical Psychology, 34(2), 330 332. Baum, L., Archer, R., Forbey, J., & Handel, R. (2009) A review of the MMPI A and the MACI with a n emphasis on juvenile justice samples. Assessment 16(4), 384 400. Becker, J.V., and Johnson, B. R. (2001) Treating Juvenile Sex Offenders. In Ashford, J. B., Sales B. D. and Reid, W. H. (Eds.) Treatment of Adult and Juvenile Offenders with Special Needs Washington D.C.: American Psychological Association Press. Blaske, D., Borduin, C., Henggeler, S., & Mann, B. (1989) Individual, family, and peer characteristics of adolescent sex offenders and assaultive offenders. Developmental Psychology 25 ( 5), 846 855. Colligan, R.C. & Using the MMPI with adolescents (pp. 215 240). Hillsdale, NJ: Lawrence Erlbaum Associat es. Craig, L., Browne, K., Beech, A., & Stringer, I. (2006) Differences in personality and risk characteristics in sex, violent, and general offenders. Criminal Behaviour and Mental Health 16, 183 194. Davis, K., & Archer, R. (2010). A critical revi ew of objective personality inventories with sex offenders. Journal of Cl i nical Psychology 66(12) 1254 1280. Gottesman, I.I., Hanson, D.R., Kroeker, T.A., & Briggs, P.F. (1987). New MMPI normative data and power transformed T score tables for the Hathaway Monachesi Minnesota Cohort of 14,019 15 year olds and 3,674 18 year olds. In R. P. Archer, Using the MMPI with adolescents (pp. 241 297). Hillsdale, NJ: Lawrence E rlbaum Associates.
39 Herkov, M., Gynther, M., Thomas, S., & Myers, W. (1996). MMPI differences between adolescent inpatients, rapists, sodomists, and sexual abusers. Journal of Personality Assessment 66(1), 81 90. Fla. Stat. § 985.18 (2010) Federal Bureau of Investigation. (2009). Uniform crime reports. Retrieved from http://www2.fbi.gov/ucr/cius2009/offenses/violent_crime/index.html Freeman, K., Dexter Mazza, E., & Hoffman, K. (2005) Comparing personality characteristics of juvenile sex of fenders and non sex offending delinquent peers: A preliminary investigation. Sexual Abuse: A Journal of Research and Treatment 17(1). Glaser, B., Calhou n G., & Petrocelli, J (2002). Personality characteristics of male juvenile offenders by adjudicated o ffenses as indicated by the MMPI A. Criminal Justice and Behavior 29 (2), 183 201. Graham, J.R. (2006). MMPI 2: Assessing personality and psychopathology. New York: Oxford University Press. Gudjonsson, G. H., & Sigurdsson, J. F. (2000). Differences and similarities between violent offenders and sex offenders. Child Abuse & Neglect, 24 (3), 363 372. Hall, G., Sh epherd, J., & Mudrak, P. (1992). MMPI taxonomies of child sexual and nonsexual offenders: A cross validation and extension. Journal of Personality Assessment 58(1), 127 137. Hathaway, S., & Monachesi, E., (1957). The personalities of predelinquent boys. Journal of Criminal Law, Criminology, and Police Science 48, 149 163. Jonson Reid, M. & Way, I. (2001) Adolescent sexual offenders : incidence of childhood maltreatment, serious emotional disturbance, and prior offenses. American Journal of Orthopsychiatry 71(1), 120 130. Langevin, R., Wright, P, & Handy, L. (1989) Characteristics of sex offenders who were sexually victimized as c hildren. Sexual Abuse: A Journal of Research and T reatment 2(3), 227 253. Losada Paisey, G. (1998). Use of the MMPI A to assess personality of juvenile male delinquents who are sex offenders and nonsex offenders. Psychological Reports 83, 115 122. Lyman, D., Charnigo, R., Moffitt, T., Raine, A., Loeber, R., & Stouthamer Loeber, M. (2009). The stability of psychopathy across adolescence. Development and Psychopathology 21, 1133 1153.
4 0 Morton, T., Farris, K., & Brenowitz, L. (2002) MMPI A scores and high points of male juvenile delinquents: Scales 4,5, and 6 as markers of juvenile delinquency. Psychological Assessment 14(3), 311 319. MMPI A forensic case studies: Uses in documented court decisions Journal of Personality Assessment 90(3), 215 226. Oxnam, P., & Vess, J. (2008) A typology of adolescent sexual offenders: Millon adolescent clinical inventory profiles, developmental factors, and offence characteristics. The Journal of Forensic Psychiatry& Psyc hology 19(2), 228 242 Parker, J., Morton, T., & Lingefelt, M., & Johnson, K. (2005) Predictors of serious and violent offending by adjudicated male adolescents. North American Journal of Psychology 7(3), 407 418. Rich, P. (2003). Understanding, assess ing, and rehabilitating juvenile sexual offenders. Hoboken, New Jersey: John Wiley & Sons.
41 BIOGRAPHICAL SKETCH Amanda Schwait was born an d raise degree in Psychology from the University of Florida in 2006. Following her undergraduate work, she worked as an English teacher for Spanish speaking students in Central America. She is currently doing her graduate studies at the University of Florida, where she is a student in the Counseling Psychology program.