1 THE EFFECTS OF ATHLETIC PARTICIPAT ION ON ADOLESCENT SEXUAL DEBUT By SALLY ANN MOORE A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2007
2 2007 Sally Ann Moore
3 To my parents, who have been right by my side every step of the way
4 ACKNOWLEDGMENTS I first recognize and thank my family, especia lly my parents. Wit hout their unconditional love and encouragement, I could not have reac hed my goals. A special thanks goes to my brother and all of my friends w ho patiently supported me through th is process. I also thank the amazing people at Tampa Gymnastics and Dance who introduced me to the world of athletics. Gymnastics, and the coaches and teammates w ho helped me along the way, has provided me with unforgettable opportunities that still shape me today. Next, I express my gratitude to the member s of my committee. Without the individual guidance and expertise offered by each of these professors, this thesis would not be possible. I am especially thankful to Dr. Rose Barnett, w hose role as committee chai r, professor, mentor, and friend has been invaluable, both personally a nd professionally. I extend my thanks to Dr. Mark Brennan for the unwavering enthusiasm, ex treme patience, and generosity with time and knowledge that helped keep me going. Las tly, I thank Dr. Heather Gibson for her encouragement, attention to detail, and friendliness. Finally, I extend a big thank you to the staff of the department of Family, Youth and Community Sciences.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES................................................................................................................ .........9 LIST OF TERMS.................................................................................................................. .........10 ABSTRACT....................................................................................................................... ............11 CHAPTER 1 INTRODUCTION..................................................................................................................13 Adolescent Sexual Behavior...................................................................................................13 Sport Participation............................................................................................................ ......14 Theory......................................................................................................................... ............15 Purpose of Study............................................................................................................... ......16 Research Questions.........................................................................................................16 Hypotheses..................................................................................................................... .17 2 MATERIALS AND METHODS...........................................................................................19 Research Design................................................................................................................ .....19 Population/Sampling Frame...................................................................................................19 Data Collection................................................................................................................ .......21 Instrumentation................................................................................................................ .......21 Athletic Participation.......................................................................................................22 Athletic Identity Measurement Scale..............................................................................23 Ecological Variables........................................................................................................24 Brief Sensation Seeking Scale (BSSS)............................................................................26 Sexual Activity................................................................................................................28 Demographics..................................................................................................................29 Data Analysis.................................................................................................................. ........30 3 LITERATURE REVIEW.......................................................................................................31 Theoretical Framework.......................................................................................................... .31 Ecological Model of A dolescent Sexual Debut......................................................................32 Individual Level...............................................................................................................33 Familial Level................................................................................................................. .37 Extrafamilial Level..........................................................................................................38 Athletic Participation......................................................................................................... .....41 Athletic Participation and Adolescent Sexual Behavior.........................................................44
6 Summary........................................................................................................................ .........49 4 RESULTS........................................................................................................................ .......50 Descriptive Results............................................................................................................ .....50 Sample Demographics.....................................................................................................50 Athlete Status................................................................................................................. ..53 Sexual Debut...................................................................................................................54 Age of Sexual Debut.......................................................................................................56 Bivariate Analyses............................................................................................................. .....56 Demographics..................................................................................................................56 Sexual Behavior...............................................................................................................58 Age of Sexual Debut................................................................................................58 Sexual Activity.........................................................................................................59 Athlete Status................................................................................................................. ..60 Multivariate Analyses.......................................................................................................... ...62 Age of Sexual Debut.......................................................................................................63 Sexual Activity................................................................................................................66 Athlete Status................................................................................................................. ..68 Summary........................................................................................................................ .........71 5 DISCUSSION..................................................................................................................... ....75 Research Questions............................................................................................................. ....75 Sexual Activity................................................................................................................ .......81 Athletic Participation.......................................................................................................82 Ecological Variables........................................................................................................83 Sensation Seeking............................................................................................................84 Ecological Model of Sexual Debut.........................................................................................85 Individual Level...............................................................................................................86 Familial Level................................................................................................................. .96 Extrafamilial Level..........................................................................................................98 Contribution to the Literature...............................................................................................103 Limitations and Delimitations..............................................................................................105 Implications for the Future...................................................................................................105 Research....................................................................................................................... .105 Practice....................................................................................................................... ...106 Public Policy..................................................................................................................108 Summary........................................................................................................................ .......109 APPENDIX A CONSENT FORM................................................................................................................110 B INSTRUMENTATION........................................................................................................111 C SCALE CORRELATION TABLES....................................................................................118
7 D TABLES OF SIGNIFICANT RESULTS.............................................................................123 E SELECT RESPONSE TO OPEN-ENDED QUESTIONS...................................................127 LIST OF REFERENCES.............................................................................................................130 BIOGRAPHICAL SKETCH.......................................................................................................139
8 LIST OF TABLES Table page 4-1 Gender frequency........................................................................................................... ....51 4-2 Race/ethnic origin frequency.............................................................................................51 4-3 Religious affiliation frequency..........................................................................................52 4-4 Christian denomi nation frequency.....................................................................................53 4-4 Athlete status frequency................................................................................................... ..53 4-5 Sexual debut frequency..................................................................................................... .54 4-6 Debut age for sexual behaviors..........................................................................................55 4-7 Cross-tabulation of leve l of contact and gender................................................................57 4-8 Cross-tabulation of ge nder and athlete status....................................................................61 4-9 Cross-tabulation of religious affiliation and athlete status................................................61 4-10 Cross-tabulation of race/et hnic origin and athlete status...................................................61 4-11 ANOVA for AIMS by athlete status..................................................................................62 4-12 Comparison of four multivariate models on age of sexual debut......................................72 4-13 Comparison of four multivaria te models on sexual activity..............................................73 4-14 Comparison of reduced multivar iate models on sexual activity........................................74
9 LIST OF FIGURES Figure page 3-1 Ecological model of selected risk factors for adolescent sexual activity...........................33
10 LIST OF TERMS Adolescence The period of the life betw een puberty and adult status. For the purpose of this study, ages 11-17 years old represent the time period of adolescence. Emerging Adulthood The period of life betw een adolescence and adulthood. For the purpose of this study, ages 18-25 years old represent the time period of emerging adulthood. Protective Factors Factors at the individual or environmental level that moderate or mediate the impact of risk factor s, making it less likely that a problem behavior will occur. Risk Factors Factors at th e individual or environmenta l level that increase an individuals vulnerability to ne gative outcomes and/or problem behaviors. Sexual Debut The first time an indi vidual engages in sexual intercourse. Voluntary Sexual Debut Choice at the time of sexual debut.
11 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science THE EFFECTS OF ATHLETIC PARTICIPAT ION ON ADOLESCENT SEXUAL DEBUT By Sally Ann Moore August 2007 Chair: Rosemary V. Barnett Major: Family, Youth and Community Sciences My research investigated the relationship between athletic partic ipation and adolescent sexual debut. The effect of athletic participa tion on overall sexual activity was also examined. An ecological perspective was used to explore an d identify various risk and protective factors that might influence sexual de cision making in adolescence. A sample of 437 undergraduate students enrolle d in general education courses completed the 70-item Athletic Participati on and Sexual Activity Questionnaire created for use in this study. The questionnaire consisted of the following concep tual areas: athletic participation, Athletic Identity Measurement Scale (A IMS), ecological variables, Brief Sensation Seeking Scale (BSSS), sexual activity, and sociodem ographics. Data were analyzed using chi-square tests, ttests, ANOVA, and multiple regression analyses. Results indicate that factors at the individual, familial, and extrafamilial level affect both age of sexual debut and overall sexual activity. Specifically, positive values, risk avoidance, school attachment, parental monitoring, and remain ing involved in sport in college were found to be protective factors, while higher propensity fo r sensation seeking and increased feelings of connectedness to school and others were related to higher sexual activity scores. Findings suggest that the link between bei ng an athlete and sexual behavior may not be as direct or as
12 simple as suggested in prior studi es. Instead, it seems that this complex relationship is mediated by the influence of various factor s related to the sport particip ation experience. Additionally, while prior research has found that adolescents engage in oral se x later than sexual intercourse, this study found that oral sex occurred at a younger age and prior to sexual intercourse. Recommendations for future research include exploring factors at th e macrosystem level, examining the effect of participation in other extr a-curricular activities, and delving further into this topic using targeted samples of specific athlete populations. Conclusions reached through this study have important implications for yout h practice and public po licy. Specifically, programs and policies designed to positively impact sexual behavior must take into account the entire environment in which teens live.
13 CHAPTER 1 INTRODUCTION David Beckham. Kobe Bryant. Anna Kourni kova. Considering the lives and actions of some of the most popular professional athletes reveals an apparent c onnection between sports and sex. Whether it is a scandalous affair or heavily marketed se x appeal, athletes often exhibit some aspect of enhanced, and even glorified, sexuality, generally without detriment to their personal life or career. This obs ervation leads one to question wh ether this relationship holds true for only the super athletes those who gain worldwide fame and millions of dollars for playing their sport, or for the more casual partic ipants as well. Specifi cally, might there be a similar relationship between being an athlete and sexual behavior among adolescents? This thesis explored this relationship, seeking to unde rstand a small part of the complicated issue of adolescent sexual debut. Adolescent Sexual Behavior In its annual Youth Risk Behavior Surveilla nce System (YRBSS) report, the CDC found that in 2003, 34% of surveyed adolescents in th e United States were currently sexually active, defined as having had sexual intercourse with on e or more people in the past three months (Grunbaum, Kann, Kinchen, Ross, Hawkins, Lowr y, et al., 2004). Additionally, 46.7% reported having had sexual intercourse at some point in their lifetime, 7.4% before the age of 13, and 14.4% reported having had four or more partners in their lifetime. These staggering numbers take on even grea ter significance when the following statistics are considered. Findings from th e same YRBSS show that only 63% of currently sexually active teens reported that they or their partner had used a condom during last sexual intercourse and just 17% reported that they or their partner had used bi rth control pills before last sexual intercourse. Of those adolescents who were currently sexual ly active, a full 25.4% had drunk alcohol or used
14 drugs before last sexual intercourse. Furtherm ore, the CDC reported that, Nationwide, 4.2% of students had been pregnant or had gotten so meone pregnant (Grunbaum et al, 2004, p. 318). It is clear that many adolescen ts are choosing to engage in sexual activity, and are dealing with the consequences of that decision. What is not quite so clear, however, is why these adolescents choose to have sexual intercourse and, conversely, why the remainder choose to abstain. A variety of factors have been sugge sted as indicators of adolescent sexual debut, including race/ethnicity (Upc hurch, Levy-Storms, Sucoff, & Aneshensel, 1998), gender (Upchurch, Levy-Storms, Sucoff, & Aneshensel, 1998), dating status (L ittle & Ranking, 2001), concern about future vocationa l opportunities (Small & Luster, 1994), personality differences (Miller, Lynam, Zimmerman, Logan, Leukefel d, & Clayton, 2004), peer crowd affiliation (La Greca, Prinstein, & Fetter, 2001), and participation in certain extracurricu lar activities (Miller, Sabo, Farrell, Barnes, & Melnick, 1998). As this significantly condensed list suggests, the ways in which teens may be influenced in their deci sion to have sex or not are vast and complex. Sport Participation One particular factor that may impact an a dolescents decisions a bout sexual activity is participation in athletics. Res earchers have reported various eff ects of being involved in sports, both positive (Kirkcaldy, Shephar d, & Siefen, 2002; Petitpas, Van Raalte, Cornelius, & Presbrey, 2004) and negative (Baumert, Henderson, & Thom pson,1998; La Greca, Prinstein, & Fetter, 2001). Although studies of adolescent sport particip ation generally consider athletes to be one population, there is some evidence to suggest that various factors within sport participation cause the experience of one athlet e to be markedly different from that of another. Differences in sport experience may then affect the ultimate effect that participation has on athletes thoughts and behaviors. Several important f actors have been identified, incl uding type of sport, level of contact, team vs. individual sport, intensity of s port participation, and level of athletic identity.
15 The sparse research designed to specifically examine athletic participation and adolescent sexual behavior is just as cont radictory. While some researcher s suggest that being an athlete may protect against sexual activity and teen pregnancy (Lehman & Koerner, 2004; Savage & Holcomb, 1999), others disagree (Dodge & Jaccard, 2002). Furthermore, it has also been suggested that the degree of pr otection against sexual activity offe red by athletic participation is different for males and females. Specifically, bein g an athlete was found to be a protective factor for girls, leading to an older age of sexual de but, but a negative factor for boys, leading to a younger age of sexual debut (Sabo, Farrell, Barn es, & Melnick, 1998). Fu rther research is clearly needed to clarify this relationship. This th esis attempts to add to the literature regarding athletic participation and adolescent sexual debut. Theory The overarching theoretical framework used in this thesis to examine the relationship between athletic participation and adolescent sexual debut is the ecological model. The original social ecological model, proposed by Bronfe nbrenner (2000; Bronfenbrenner & Morris, 1998), was an effort to explain the multifaceted environm ent in which children and adolescents live and make decisions. This model, emphasizing the impor tance of social interactions, postulated that factors influencing child and adolescent development exist within multiple systems, or levels, which make up an individuals social ecology. Bronf enbrenners theory stresses the fact that in the process of developing as a person, children and adolescents ar e impacted by a wide variety of experiences that cannot be wholly teased apart. This idea has provided th e basis for theoretical perspectives exploring many concepts a nd behaviors, including sexual debut. Specifically, Small and Luster (1994) designe d a research-based, ecological, risk-factor approach to examine adolescen t sexual activity. This appr oach seeks to understand sexual behavior, recognizing that indivi duals do not make decisions without being influenced by the
16 world around them, and that various factors work to impact thought s, feelings, and behaviors. Small and Luster posit three levels of influence, with risk factors at each level: individual (e.g., drug use, intellectual ability), familial (e.g., fam ily structure, parental monitoring), and extrafamilial (e.g., school experiences, peer influence). Involvement in sports, an extra-curricular activity, falls into the extra-familial sphere. In addition, being an athlet e may affect the other levels. This research, then, hopes to determine th e relationships between an individuals social ecology, athletic participation, and timing of sexual debut. Purpose of Study The purpose of this study was to inves tigate the relationship between athletic participation and adolescent sexual debut. A lthough numerous studies have examined the behavioral and psychological char acteristics of athletes as comp ared to the general population, and many more have set out to identify factors le ading to or protecting ag ainst adolescent sexual debut, very few have tried to forge a connecti on between the two areas. This study aims to bridge that gap. Research Questions The primary research questions examined in this study are: Q1. Does being an athlete af fect timing of sexual debut? Q1.1. Does intensity of participat ion affect timing of sexual debut? Q1.2. Does level of athletic identity affect timing of sexual debut? Q2. Does type of sport affe ct timing of sexual debut? Q2.1. Does level of contact in a sport affect timing of sexual debut? Q2.2. Does team vs. individual sp ort affect timing of sexual debut? Q3. Do ecological variables af fect timing of sexual debut? Q3.1. Does propensity for sensation s eeking affect timi ng of sexual debut?
17 Q4. Do ecological variables differ between athletes and nonathletes? Q4.1. Does propensity for sensation s eeking differ between athletes and nonathletes? Hypotheses The primary hypotheses examined in this study are: Q1. HO: Being an athlete does not affect timing of sexual debut. HA: Being an athlete does affe ct timing of sexual debut. Q1.1. HO: Intensity does not affect timing of sexual debut. HA: Intensity does affect timing of sexual debut. Q1.2. HO: Level of athletic identity does not affect timing of sexual debut. HA: Level of athletic identity doe s affect timing of sexual debut. Q2. HO: Type of sport does not affe ct timing of sexual debut. HA: Type of sport does affect timing of sexual debut. Q2.1. HO: Level of contact does not a ffect timing of sexual debut. HA: Level of contact does aff ect timing of sexual debut. Q2.2. HO: Team vs. individual sport does not affect timing of sexual debut. HA: Team vs. individual sport does affect timing of sexual debut. Q3. HO: Ecological variables do not a ffect timing of sexual debut. HA: Ecological variables do a ffect timing of sexual debu t. Q3.1. HO: Propensity for sensation seeking does not affect timing of sexual debut. HA: Propensity for sensation seeking does affect timing of sexual debut. Q4. HO: Ecological variables do not differ between athletes and nonathletes. HA: Ecological variables do differ be tween athletes and nonathletes. Q4.1. HO: Propensity for sensation seeking doe s not differ between athletes and
18 athletes. HA: Propensity for sensation seeking doe s differ between athletes and nonathletes.
19 CHAPTER 2 MATERIALS AND METHODS Research Design This study employed a cross-sectional resear ch design. According to de Vaus (2004), Cross-sectional designs have th ree distinctive features: no time dimension; reliance on existing differences rather than change following interv ention; and groups based on existing differences rather than random allocation (p. 170). Thus, data collection occurred at one point in time (spring semester 2007), and examined a wide range of cases. Groups were not randomly allocated, but were rater assigned based on prior athletic participation, an existing difference that did not involve researcher intervention. Cro ss-sectional designs permit the researcher to examine differences between groups by looking at multiple variables at the same time. Crosssectionals are a commonly used research desi gn, particularly for studi es in which certain variables cannot be manipulated for ethical reas ons (Bryman, 2004), and are also a good choice for exploratory research studies. Although studi es using a cross-sectional design look for relationships between two or more variables, th ey cannot conclusively ex plain the direction of the relationship. Correlations can be stated, but causation cannot be established. For this study, the predictor variables (a thletic participation, eco logical variables, sensation seeking) and the outcome variables (age of sexual debut, overall sexual activity) were examined to determine whether any relationships were present. Population/Sampling Frame The theoretical population for this study cons isted of students enrolled at a large southeastern university in th e spring of 2007. The population was comprised entirely of undergraduate students. The sample population fo r this study consisted of the approximately 36,000 undergraduate students enrolled at this un iversity during the target semester.
20 Undergraduates, particularly freshmen and sophom ores, were targeted because these individuals are not far removed from adoles cence and, thus, were thought to be more likely to accurately recall events from this time in their lives. Us ing the criteria established by Krjcie and Morgan (1970), it was determined that a sample size of approximately 400 would be necessary to ensure a representative sample. According to resear ch conducted by the Cent ers for Disease Control (CDC), approximately half of a dolescents in the United States re port being involved in at least one sport (Grunbaum et al., 2004) The researcher knew of no compelling reasons why students at this University might differ from other student s; thus, it was expected that roughly half of the participants in this study would have been athletes during high school. Required general education cour ses enrolling primarily freshm en and sophomore students were randomly selected and professors were contacted by email to ask permission for the researcher to administer the instrument. A comp rehensive list of all ge neral education classes was obtained and those courses not offered during Spring semester 2007 were removed. A random sample was gathered by choosing every sevent h course. Each instructor for the selected courses was emailed to inform them of the study and to ask permission for their class to participate. This first email contact was follo wed by a telephone call if there was no response. Following the initial email and tele phone call, a second email was sent A total of 76 instructors were initially contacted. Fi fty-seven instructors did not respond to any communication, 16 declined participation, a nd 3 agreed to allow data collection in their classes. The three classes sampled for data collection came from two of the largest colle ges within the university and represent wide variation in s ubject area (communications, social science, nutrition), suggesting that a broad cross-section of respondents wa s sampled. The researcher stopped contacting instructors and collecting data when the goal sample size was reached.
21 Data Collection For each undergraduate class, the researcher c oordinated a time with the professor to come to class and collect data Every student in attendance was as ked to participate. All potential participants were informed th at their participation was volunt ary, confidential, and anonymous, in accordance with the universitys Internal Review Board (IRB) regulations. Students were instructed not to participate if they had already completed the que stionnaire in anot her class, so that there would be no duplicate participants. A total of 437 participants fr om three separate classes comp leted the questionnaire for this study. Very few (n = 6) students present in clas s at the time of data collection chose not to participate in the study. In all of these cases, students offered reasons for their nonparticipation that were unrelated to the study itself (e.g., I need to use this available time to study for an exam.). It was, therefore, assumed that t hose students who chose not to complete the questionnaire were not significantl y different from those who did pa rticipate in the study. Of the completed questionnaires, 417 were used in data analysis. Two cases were excluded due to participants reporting being younger than eighteen years old. Eight cases were excluded due to participants reporting ages over 25, meaning they are no longer c onsidered part of the emerging adulthood age range. Ten cases were excluded because respondents indicated that their sexual debut was not voluntary. Instrumentation The 70-item Athletic Particip ation and Sexual Activity Ques tionnaire was created using a modified total design method (Dillman, 2000) for use in this study to explore the research questions set out in Chapter One (see Appe ndix B). The complete questionnaire took approximately 20 minutes to complete. The quest ionnaire consisted of the following conceptual
22 areas: athletic participation, Athl etic Identity Measurement Scal e (AIMS), ecological variables, Brief Sensation Seeking Scale (BSSS), se xual activity, and sociodemographics. Athletic Participation This section included a multiple-choice item aski ng participants to classify themselves as one of the following choices: a.) a high school athl ete who is now a formal athlete competing for the university; b.) a high school athlete who is now participating in univers ity intramural or club sports; c.) a high school athlete who no longer participates in sports; or d.) a high school nonathlete. Respondents who indicated that they were not athletes during high school were asked to skip the remainder of this section, as we ll as the Athletic Identity Measurement Scale. Respondents who indicated that th ey were athletes during high school were asked to complete the remaining questions in this section. Respondents were asked which sports they par ticipated in and were given the opportunity to indicate a primary sport in the ev ent that they participated in more than one sport. Each of the sports reported by participants as a primary s port was assigned a Level of Contact code (1 = Noncontact, 2 = Contact, 3 = Collision). These classifications were made following a coding system created by Silva (1983) for his research on type of sport and rule violating behavior. Many of the sports included in this study were also part of Silvas re search and, thus, were previously categorized. For these cases, the leve l of contact assigned to the sport by Silva was also used for this study. Those sports not incl uded in Silvas study were assigned to categories based on the description of collision, contact, a nd noncontact sports included in the variable coding section of his pub lication. Specifically, Categorization of a sport into noncontact, c ontact, or collision groupings was based upon the degree that physical contact is an implicit (as opposed to in cidental) part of appropriate player behavior. For example, in baseball physical contact at any base between a sliding base runner and the defensive player is incide ntal, not required, and extremely well defined in the formal rule structure. Sports such as basketball and soccer allow considerable
23 physical contact between opponents; however, extreme contact or collisions are not implicit or required actions of these sports. Football, ice hockey, lacrosse (mens), and rugby are all sports that have contact and collision behavi ors as integral predesigned aspects of appropriate goal-d irected behavior (p. 442). Each sport was also assigned a Team vs. Indivi dual Sport code (1 = Team, 2 = Individual). These coded groupings were used to create the Level of Contact and Team vs. Individual Sport variables used for analysis. Respondents were also asked how many hours per week, on average, they participated in their spor t; this item was used as an in dicator of sport participation. Athletic Identity Measurement Scale As previously mentioned, participants who re ported not participating in sport while in high school did not complete this section. The Athl etic Identity Measurement Scale (AIMS) is a seven-item scale designed to assess athletic identity in both men and women. Research has indicated that this scale is a multidimensional model, with three highly correlated first order factors (Social Identit y, Exclusivity, and Negativ e Affectivity) which are subordinate to a higher order athletic identity factor. The first three items (I considered myself an athlete; I had many goals related to sport; and Most of my friends we re athletes) are indicators of Social Identity; the following two items (Sport was the most important part of my life; and I spent more time thinking about sport than anything else) are indicators of Exclus ivity; and the last two items (I felt bad about myself when I did poorly in sport; and I would have been very depressed if I were injured and could not compete in sport) are indicators of Negativ e Affectivity. The scale, as developed, includes all of these items in the pres ent tense. Because this study asked respondents to think back to a previous time in their life, how ever, the researcher changed all items to the past tense. Individual items are rated on 7-point Like rt-type scales and then summed for an overall athletic identity score. Thus, the AIMS is c onsistent with a conceptualization of athletic identity as a superordinate construct incorpora ting disparate aspects of sport-specific self-
24 identity (Brewer & Cornelius, 2001, p. 104). For th e purpose of this study, the researcher added an additional item to assess pe rception of membership in the popular group in high school. Though participants answered this question in the same manner as they answered the other items in this section, responses were c onsidered separately and not include d in the total AIMS score. The AIMS has been used and tested in many other research studies (Brown & Hartley, 1998; Grove, Fish, & Eklund, 2004; Horton & Mac k, 2000; Lantz & Schroeder, 1999; Lau, Fox, & Cheung, 2005; Tasiemski, Kennedy, Gardner, & Blaikley, 2004). It has been found to be effective for, and discriminating across, differe nt levels of sport involvement (nonathlete, recreational athlete, competitive athlete) and has not been found to be significantly correlated with conceptually dissimilar cons tructs, such as social desirabi lity, self-esteem, and self-rated sports competence (Brewer, Van Raalte, & Lind er, 1993). General sta tistical support for the psychometric integrity of the AIMS has b een obtained (Brewer & Cornelius, 2001). Analyses using data gathered for this study support the us e of this scale. Pearson Correlations (Table C-1) were run for the to tal AIMS scale and each statement and it was determined that each individual item was highly corre lated to the total AIMS scale, as well as to all other individual items. To ensure a sufficien t level of internal cons istency, Cronbachs alpha coefficients were conducted for al l scales included in this study. Cronbachs alpha for the total AIMS scale was .89. Additionally, Pearson Correl ations (Tables C-2, Table C-3, and Table C-4) were run for each of the three athletic identity domains, with similarly positive results. Cronbachs alpha for the Social Identity, Excl usivity, and Negative Affectivity scales was .80, .88, and .76, respectively. Ecological Variables This section asked questions about the participants social ecology. It consisted of 32 items rated on a 5-point Likert-type scale measur ing variables at the individual, familial, and
25 extra-familial level identified in the literature review as being important to timing of sexual debut. Response options ranged from 1 (strongly agre e) to 5 (strongly disagree). For analysis purposes, all but one of these item s were reverse-coded so that a higher score reflected more agreement. The remaining item (My close friend s starting having sex when they were in high school) was not reverse-coded, since more agreemen t with this statement reflects a risk factor, while more agreement with the other items include d in the same scale does not reflect risk. An exploratory factor analysis conducted on this gr oup of items led to the formation of six distinct factors, each with its own composite score. Data resulting from this survey were factor analyzed using principal axis f actoring with a varimax rotation (K im & Mueller, 1978). Prior to selection of factor items, the following criteria were establis hed: a factor loading of .35 or higher; at least a .10 difference between the items lo ading with its factors; and interpretability. A review of the factors with eige nvalues greater than 1.0 and furthe r analysis of scree test plots suggested that a four, five, or si x factor solution would be most a ppropriate, as the scree test had distinct and obvious breaks at th ese points. Further considerati on of the items included within each of these factors indicated that the six f actor solution made the most theoretical and conceptual sense (Kim & Mueller, 1978). Pear son correlations for each of the six indices supported these groupings (Tables C-5, C-6, C-7, C-8, C-9, and C-10). Two items (I felt like I received a good education; and My school work was very importa nt to me) did not meet the established criteria for inclusion in to any of the six factors. They were, therefore, considered as individual items for analysis. Following is a summary of the final six factors: Risk Avoidance : Included were the statements I avoided smoking cigarettes; I avoided using marijuana; I avoided drinking alcohol; My close friends avoi ded smoking cigarettes; My close friends avoided using marijuana; My close friends avoided drinking alcohol; and My close friends starting having sex when th ey were in high school. Cronbachs alpha was .85, suggesting a sufficient level of relia bility.
26 Parental Closeness : Included were the statements I felt cl ose to at least one parent; I talked to at least one parent about sex; My parents always supported me; I felt like I could talk to my mother about anything; I felt like I could ta lk to my father about anything; My mother was always there for me; and My father was always there for me. Cronbachs alpha was .80. Parental Monitoring : Included were the statements My parents always knew where I was going when I left home; My parents always knew who I was with; and My parents always knew were I was after school Cronbachs alpha was .87. Connection to School and Others : Included were the statem ents I liked going to high school; High school sport was very important in my life; I was involved in extracurricular activities other than sports at my high school; I had lots of friends at my high school; I had at least one adult other than my parents who really cared about me when I was in high school; and People in my neighborhood watched out for me. Cronbachs alpha was .58. This is a borderline alpha leve l, as a general cutoff of .60 is usually used. Despite the low level of reliability, the researcher opted to group these items into a single index, rather than considering them individually. This decision was based on a de sire to keep the number of variables to a manageable number for the multivariate analyses and to preserve the degrees of freedom. Additionally, this decision was supported by the fact that multiple factor analysis runs continued to group this set of statements together in this manner. Values : Included were the statements I attended religious services regularly; Religion was very important in my life; and My parents would have strong ly disapproved of me having sex while in high school. Cronbachs alpha was .79. Popularity : Included were the statements I tried to do things to be more popular; If my friends were doing something, I wanted to do it too; At my school, it was considered cool to have sex; and I cared about high school mostly for the social atmosphere. Cronbachs alpha was .58. This is a borderline alpha scor e when compared to the usual cutoff of .60. Similarly to the Connection to School and Ot hers scale described above, the researcher chose to group these items into a single index despite the low level of reliability. This grouping was maintained in an effort to keep the number of variab les to a manageable number for the multivariate analyses and to preserve the degrees of freedom. Furthermore, multiple factor analysis runs continued to gr oup this set of statements together in this manner. Brief Sensation Seeking Scale (BSSS) The Brief Sensation Seeking Scale (BSSS) is a self-report Likert-type scale adapted from Form V of the Sensation Seeking Scale (Zucke rman, 1968) and a group of items taken from the SSS-V and altered for use with adolescents (Huba et al., 1981). The SSS-V is the typical assessment for sensation seeking, and has been wi dely and successfully used for a variety of
27 research studies. It is, however, for many reasons including several that are significant to this study, not ideal for many types of common research contexts. The SSS-V consists of 40 items in forced-choice format. This large quantity of items makes it unwieldy for already lengthy surveys or for questionnaires designed to be administered within restricted time periods. Additionally, the forced choice format is cumbersome and ma y be particularly difficult for adolescent participants. Furthermore, many terms or phrases intended to reflect well-known or familiar ideas are no longer meaningful or pertinent to ad olescents or young adults. To address these and other limitations, Hoyle et al. ( 2002) created the BSSS to be suita ble for survey research with adolescents and young adults. Following this same logic, the BSSS was thought to be the most appropriate measure of sensation se eking for the purposes of this study. The BSSS consists of 8 items. Two items ar e included to represent each of the four primary dimensions of sensation seeking (Expe rience Seeking, Boredom Susceptibility, Thrill and Adventure Seeking, and Disinhibition). Indivi dual items are rated on a 5-point scale labeled strongly disagree, disagree, neither disagr ee nor agree, agree, and strongly agree and are summed for a total BSSS score. For purposes of the multivariate analyses, each of these items was reverse-coded so that a higher score reflected more agreement. The BSSS has been found to have strong reliability and validity, with solid psychometric characteristics that hold up across sex, age, and ethnic categories (Hoyle et al., 2002, p. 411). The construct validity of this scale has been further supported by several additional studies (Donohew, Zimmerman, Cupp, Novak, Colon, & Abell, 20 00; Palmgreen, Donohew, Lorch, Hoyle, & Stephenson, 2001; Stephenson, Palmgreen, Hoyle, Donohew, Lorch, & Colon, 1999). Analyses using data gathered in this study support the use of this scale. Pearson Correlations (Table C-11)
28 were conducted and it was determined that each individual item was highly correlated to the total BSSS scale, as well as to all other in dividual items. Cronbachs alpha was .77. Sexual Activity This section asked questions about the participants se xual activity. For each of a series of sexual behaviors, respondents were asked to indicate on 5-point Likert-type scale ranging from Never to A lot the frequency with which th ey have engaged in the behavior. If they indicated that they had ever engaged in the beha vior, they were asked to provide the month and year in which they first engaged in the behavior In order, the behavi ors are kissing; French kissing; touching a partners breast or having your breast touched by a partner; touching a partners penis or having your pe nis touched by a partne r; touching a partner s vagina or having your vagina touched by a partner; performing oral sex; receiving oral sex; and sexual intercourse. These behaviors were taken from the sexual timetable described in research conducted by Feldman, Turner, and Araujo (1999). If participants reported having had sexual intercourse, they were asked if their first time was consensual with the options to select 1. Yes or 2. No. The definition of sexual debut provided in Chapter One states that virginity loss is voluntary. Thus, the data of participants who responded No was not analyzed with the rest of the data because their sexual debut was not a choice. The last question in this section was an open-ended it em asking participants to think about what influenced their decision to have se x or to delay sex. A box with several lines was provided for them to write their answers. For purposes of analysis, the eight items measuring frequency of engaging in sexual behaviors were combined to form an index measuring level of sexual activity. The sexual activity scale was created to assess respondents overall level of sexual activity, as opposed to the single moment captured by the age of sexual debut item. Scor es on this scale, therefore,
29 represent the many sexual choi ces faced by respondents throu ghout their lifetime. Clearly, contexts and situations, and ofte n beliefs, change over time. Fa ctors important to sexual decision making in early adolescence are likely very diffe rent from those influencing sexual decisions made by college students. Consideration of to tal sexual activity might better represent these changes in context, whereas age of sexual debut can only reflect the context at that time. The decision to treat these items in this manner was informed by a series of statistical tests. Pearson Correlations (Table C-12) were run and it was determin ed that the first seven items (kissing; French kissing; touchi ng a partners breast or having your breast touched by a partner; touching a partners penis or ha ving your penis touched by a partne r; touching a partners vagina or having your vagina touched by a partner; performing oral sex; and receiving oral sex) were highly correlated with the final it em, sexual intercourse. A factor analysis was then completed, following the same procedure detailed above for th e ecological variables. This suggested that these items were all highly interrelated and were consistently grouped in a one factor solution. The conclusions reached through this battery of tests suggest that combining these eight items into a single scale was most appropriate. Cronbachs alpha for this scale was 0.95. Demographics The demographics section contained items measuring gender, race/ ethnicity, religious affiliation, and age of participants. Demographic information is collected in most research and provides important basic information for the exam ination of differences and similarity between and within populations. Respondent s were first asked to indicate their gender, with the following choices: 1. Male; and 2. Female. Th ey were then asked to indicate what they considered to be their race/ethnic origin, with the following choices: 1. Hispan ic; 2. Asian/Pacific Islander; 3. White (non-Hispanic); 4. Black (non-Hispanic); and 5. Other. Participants circling the Other choice were provided a line on wh ich they could write in their race/ethnic origin. Respondents
30 were then asked their current age, in years and m onths. Lastly, they were asked to indicate their religious affiliation, with the following choices: 1. Jewish; 2. Muslim; 3. Christian; 4. Other; and 5. No affiliation. Participants reporting their re ligious affiliation as Christian were provided a line on which they were asked to indicate to which specific Christian denomination they belonged. Likewise, those particip ants circling the Other choices were provided a line on which they could write in their religious affiliation. Data Analysis The researcher used Statistical Package for So cial Scientists (SPSS) 15.0 for all statistical tests. Data collected following these methods we re analyzed using a vari ety of tests, including descriptive statistics, bivariate analyses to ex amine possible relationships between the dependent variables and both independent and demographic va riables, and multiple regressions to explore these relationships in a multivariate context.
31 CHAPTER 3 LITERATURE REVIEW The decision of if and when to have sex for the first time is a complicated one, influenced by the presence and interaction of many factors. Although a complete identification and understanding of these factors con tinues to pose a challenge for rese archers, the rela ted literature reveals significant and applicab le findings to this end. Theoretical Framework Upon a thorough examination of many relevant studies utilizing various theoretical frameworks, Urie Bronfenbrenners ecological theory of human development emerged as a particularly well-suited approach to this re search. Bronfenbrenner (2000; Bronfenbrenner & Morris, 1998) advocated for a view of child de velopment that strongly emphasized the role of social contexts. Specifically, he proposed that there are five environmental systems and that variables in each of these system s interact to impact developmen t. The five systems in the ecological theory are (Santrock, 2001): Microsystem This is the setting in which an individual lives and in cludes the persons family, peers, school, and neighborhood. Most of the direct interactions with various social agents (e.g., parents, peers, and teache rs) take place within this system. Mesosystem Relations between microsystems or connections betw een contexts occur within the mesosystem. Examples are the relatio n of family experiences to school experiences, school experiences to work experiences, and fami ly experiences to peer experiences (Santrock, 2001, p. 48). From an applied perspective, thos e who work directly with youth stress the importance of observing behavior in multiple settings in order to obtain a more complete view of adolescent development.
32 Exosystem This system is involved when e xperiences in a separate or remote social setting, one in which the adoles cent does not have an active ro le, influences experiences in another more immediate context. For example, if a close teacher beco mes pregnant and takes maternity leave, the adolescents feelings toward school may change. Another exosystem is the city government, which controls pa rks, recreation centers and library facilitie s. The quality of these services impacts the lives of children and adolescents. Macrosystem The culture (the norms of be havior, beliefs, and a ll other products of a specific group that are passed on from one genera tion to the next) in wh ich adolescents live is involved in the macrosystem. Co mparing one culture to another can provide important insights into the generality of adolescent development. Chronosystem This system, involves the patterning of environmental events and transitions over the life course and sociohistorical circumstan ces (Santrock, 2001, p. 49). An example of the effects of the chronosystem is the increased acceptance of women in the workforce now as compared to 30 years ago, a phenomenon which has definite meaning for female adolescents today. Ecological Model of Adolescent Sexual Debut In an attempt to better unders tand why some teens are sexually active and others are not, Small and Luster (1994) proposed an ecological, risk-factor m odel, based on Bronfenbrenners work. An important tenet of this model is the idea of cumulative risk, which suggests that the probability of becoming sexually active increases w ith increased exposure to risk factors. The model also proposes that risks encountered by adolescents can be grouped into four categories that represent different parts of the adolescents social ecology: (a) personal factors, such as intellectual ability; (b) fa milial factors, such as socioeconomic status; (c) extra-familial factors, such as school attachment; and (d) macrosyste ms, such as media influence. The authors
33 supported their model with the iden tification of risk factors at th e individual, familial, and extrafamilial levels for both males and females. A gr aphic model presented by Small and Luster is included below (see Figure 1). Further studies have corroborated the model by demonstrating its efficacy for multiple ethnic groups (Manda ra, Murray, & Bangi, 2003; Perkins, Luster, Villarruel, & Small; 1998). This study emphasized factors within the individual, familial and extra-familial levels. Figure 3-1. Ecological model of selected risk factors for adolescent sexual activity. Small, S. & Luster, T. (1994). Adolescent sexual activit y: An ecological, risk-factor approach. Journal of Marriage and the Family, 56 181-192. Individual Level Demographic variables such as gender and ethni city have been shown to have an effect on age of sexual debut. Numerous studies suggest that males initiate sex at a significantly younger age, are affected by different ri sk and protective factors, and th ink about sex in different ways
34 than do females (Little & Rankin, 2001; Meschk e, Zweig, Barber, & Eccles, 2000; Michels, Kropp, Eyre, & Halpern-Felsher, 2005; Miller Sabo, Ferrell, Barnes, & Melnick, 1998). Similar differences in age at first intercourse and context of decisi on-making processes have been found among ethnic groups (Miller, Farrell, Barnes, Melnick, & Sabo, 2005; Perkins, Luster, Villarruel, & Small, 1998 ; Upchurch, Levy-Storms, Sucoff, & Aneshensel, 1998). A younger age of pubertal timing compared to peers has been associated with earlier timing of first sexual intercourse for both boys (Crockett, Bingham, Chopa k, & Vicary, 1996) and girls (Meschke, Zweig, Barber, & Eccles, 2000). One important individual factor that has been shown to affect adolescent sexual activity is substance use. Adolescents who use cigarett es and alcohol, and especially those who use marijuana, cocaine, or other illicit drugs, are more likely to be sexually experienced (Lowry, Holtzman, Truman, Kann, Collins, & Kolbe, 1994) These authors also suggest that the likelihood of sexual risk behaviors, such as havi ng had sexual intercourse with four or more partners and not having used condoms at last se xual intercourse, increases with the progression from alcohol and cigarettes to marijuana, coca ine, and other illicit drug s. In a recent study comparing adolescents undergoing treatment for s ubstance abuse or dependence and a sample of comparable nontreatment peers in the same comm unity, those adolescents treated for substance problems were found to initiate se xual intercourse at an earlier age, have more sexual partners, and use condoms less consistently (Tapert, Aa rons, Sedlar, & Brown, 2001) Illicit drug use in early adolescence has been found to predict fu ture sexual risk beha vior, including having multiple sexual partners, inconsistent condom use, and pregnancy (Brook, Adams, Balka, Whiteman, Zhang, & Sugerman, 2004). Additionall y, binge drinking has b een identified as a risk factor for a variety of ris ky sexual practices, including sex without the use of a condom, sex
35 with unknown partners, sex with persons who use and/or shoot-up drugs, sex with individuals whose STI status is unknown, and sex while self or partner was using, or was high on, drugs or alcohol (Langer, Warheit, & McDonald, 2001). The association between substance use and sexual activity is not limited to sexual intercourse In a study designed to determine the risky sexual practices of high school virgins, res pondents who reported usi ng cigarettes, alcohol, marijuana, and other drugs in the past year we re more likely to have engaged in higher risk sexual activities (Schuster, Bell, & Kanouse, 1996). Further, such substance use behaviors were more common among the nonvirgins than the virgins. Psychological factors also play an important pa rt in an adolescents decision to have sex. Longmore, Manning, Giordano, and Rudolph (2004) suggest that de pressive symptoms, a risk factor, are more important than self-esteem, a protective factor, in understanding sexual onset. Additionally, level of cognitive development affects an adolescents understanding of important ideas associated with sex and l ove. Tremblay and Frigon (2004) found that most of the girls in their sample (93%) reported that it is okay to ha ve a one-night stand, but also agreed that one should not have sex only for fun ( 78.5%) or if not in love (87.7%). The authors suggest that, This contradiction can be unders tood as an indication of teen s difficulties in understanding such concepts as love and intimacy, since they ar e in a period of their development during which, quite rightly, the goal is to learn about interperso nal relationships (p. 35). Perceived goals and values were particularly important for girls when considering whet her or not to initiate sex for the first time (Michels, Kropp, Ey re, & Halpern-Felsher, 2005). Multiple dimensions of individual personality have been shown to be important in understanding why some individuals engage in various sexual behavi ors. In a study utilizing the Five Factor Model of personality, Miller, L ynam, Zimmerman, Logan, Leukefeld, and Clayton
36 (2004) found that higher levels of Extraversion, Openness to Experience, and Antagonism were related to increased sexual risk-taking, while higher levels of Conscientiou sness were related to a decrease in such behaviors. For both males and females, desire for earlier autonomy, lower levels of restraint, and percepti ons of greater physical maturity relative to their peers were found to be associated with earlier ages of initia tion of sexual intercourse (Rosenthal, Smith, & de Visser, 1999). In addition, high sensation seeking scores were found to be related to risky sexual behavior in a recent study by Greene, Krcmar, Walters, Ru bin, and Hale (2000). Generally, those with higher sensation-seeking scores were likely to have a greater risk-t aking personality. Sensation seeking is a consiste nt personality characteristic, defined by Zuckerman as, the need for varied, novel, and complex sensations and experiences and th e willingness to take physical and social risks for the sake of such e xperiences (1979, p. 10). Numerous other studies have also demonstrated a link between propens ity for sensation seeking and risky sexual behavior in adolescents (Bra dy & Donenberg, 2006; Spitalkni ck, Diclemente, Wingood, Crosby, Milhausen, Sales, et al., 2007; Stanton, Li, Cottrell, & Kaljee, 2001). Greene et al. (2000) found that th e negative effect of sensation seeking was especially true for individuals who also scored high on the personal fable (a developmental construct) dimension. The authors describe personal fa ble as an expression of egocentrism often experienced in adolescence, where new a abilit y to think about thoughts leads to a fascination with ones own thoughts which are surely different from the thoughts of others, and thus a belief in ones own uniqueness and invulnerability (p 442). This percepti on of being unique may influence adolescents attitudes toward risk behavi ors, while feeling invulnerable to risk may make engaging in such behaviors seem less dangerous.
37 Familial Level Family characteristics and qualities play an important part in the sexual behavior of adolescents. Low socioeconomic status has been shown to be positively associated with earlier sexual debut and riskier sexual behavior in num erous studies (Corcoran, 1999; Mandara, Murray, & Bangi, 2003; Small & Luster, 1994). Upchurc h, Levy-Storms, Sucoff, and Aneshensel (1998) found that family structure was strongly associ ated with youths risk of sexual activity, even when ethnicity, gender, and family socioeconomic status were taken in to account. Specifically, adolescents living with both biol ogical parents reporte d a later median age at first intercourse than adolescents in any other family situation. Pa rt of this effect may be due to the increased supervision provided by the presence of two parents in the home. This possible link is supported by the finding that increased time spent alone at home was associated with an increased likelihood of adolescent sexual activity (Perki ns, Luster, Villarruel, & Small, 1998). Not only the structure of the family, but the qua lity of the relationships within the family, are significant when considering adolescent sexu al activity. Miller, Sabo, Farrell, Barnes, and Melnick (1998) concluded that family cohesi on was clearly a consideration in the sexual behavior of adolescents, significantly reducing fre quency of intercourse and number of partners, and increasing age of first intercourse (p. 117) The authors go on to say that the findings suggest a causal link, as family cohesion was meas ured several years before reported frequency of intercourse. In addition, an adolescent females positive relationship with her father has been suggested as a protective factor against teen pregnancy (Dodge & Jaccard, 2002). Similarly, Rodgers (1999) suggested that fathers, in particul ar, play an important and direct role in helping their daughters make healthy and safe decisions about their bodies and their involvement in sexual behaviors. The same study found that sexua lly active sons lacking a supportive parental
38 relationship may be more likely to reject or ignore informa tion or concerns expressed by a parent. Parental monitoring has been found to be pa rticularly important. Rodgers (1999) found that, Sexually active adolescent males and fe males who have parents who monitor them (behavioral control) are more likely to minimize their sexual risks than their peers who have parents who monitor them less (p. 106). Over all, she concludes that parental monitoring, autonomy giving, and parental communication with in the context of a supportive relationship serve as positive influences in the lives of sexually active adol escents. Permissive parental values regarding adolescent sexual behavior although more commonly reported by males than females, are a strong risk factor for bo th males and females (Small & Luster, 1994). Extrafamilial Level All of an individuals life experi ences are affected by the larger society in which he or she lives. One particularly important influence, pa rticularly for adolescents, is the mass media. Recent research has suggested a link between the mass media and the sexual attitudes and behaviors of American adolescents. For example, in their examination of the mass media as a context for adolescents sexual behavior, LEngl e, Brown, and Kenneavy (2006) concluded that, Adolescents who are exposed to more sexual co ntent in the media and who perceive greater support from the media for teen sexual behavior, report greater intentions to engage in sexual intercourse and more sexual activity (p. 186). In addition, Ward and Friedman (2006) found that more frequent viewing of talk shows a nd of sexy prime-time programs and stronger identification with popular TV characters were associated with greate r levels of sexual experience. More frequent viewing and viewi ng TV more intently for companionship were associated with greater endorsement of sexual stereotypes.
39 Another important environment in the lives of adolescents is the neighborhood. Lack of neighborhood monitoring and living in an econo mically distressed neighborhood have both been found to be related to adoles cent sexual activity (Mandara, Murray, & Bangi, 2003; Small & Luster, 1994). In a recent study of risk-factors for teenage pregnancy, Miller (2002) found that along with family SES and various parental factors, living in a dangerous/disorganized neighborhood places adolescents at risk for pregnancy. The role of peers becomes incr easingly important in adolescence. Youth spend much of their time with age cohort peers and are influen ced by them when making decisions, particularly about behaviors such as sex. In their study of adolescent male sexuality, Smith and Guthrie (2005) found that the interpersona l effects of peer affiliation exerted a strong effect on the construct of sexuality, defined by the authors as the cognitive generaliz ations about the sexual aspects of ones self that guides and shapes sexual behaviors (p. 125). Additionally, adolescents classified as sexually inexperienced reported greater associat ion with high-achieving peers than did those who initiated sex earlier (Meschke, Zwieg, Barber, & Eccles, 2000). When deciding whether or not to have sex, a dolescents take possibl e social risks into consideration (Michels, Kropp, Eyre, & HalpernFelsher, 2005). Such risks might include getting caught, losing a relationship, earning a negative reputation, and, for boys, concern about sexual performance. Although both males and fema les take the opinions of their close friends into serious consideration in matters of sexual decision-mak ing, the point of view they emphasize appears to be different (Little & Ra nkin, 2001). For males, the belief about how many of their friends are having se x is most important. For females, it is how they think their friends would feel if the friends found out they were having sex. The authors suggest that while girls initiate sex to a ttain approval and love in an intimate, dating relationship, boys do so as a
40 way to seek status. Popularity has been shown to be less important to sexually inexperienced adolescents than it is to early sexual initia tors (Meschke, Zwieg, Ba rber, & Eccles, 2000). Adolescents spend much of their time at sc hool, making this environment an important factor in adolescent decision-maki ng and behavior. Adolescents cl assified as demonstrating high educational risk (defined as pr acticing three or more educationa l risk factors including skipping class, drinking or marijuana use before or dur ing school hours, missing sc hool due to substance use, and the purchase of alcohol or other drugs on school grounds) have been shown to be more likely to engage in other health risk behavior s (Gruber & Machamer, 2000). Specifically, this study found that higher educational risk increase d the potential of being sexually active by about two times. Mancini and Huebner (2004) found that positive school attachment and school success were associated with less overall risk behavior, including becoming sexually active. In addition to school, many adolescents are invol ved in social exchanges through activities such as clubs, sports teams, or religious groups These extracurricular outlets have been shown in many cases to have a positive, protective e ffect on sexual debut. Crockett, Bingham, Chopak, and Vicary (1996) found that, for girls, regular church attendance was as sociated with later sexual initiation, suggesting that conventional a ttachments play an important role in sexual decision-making. Adolescents with strong religious values also tended to have significantly lower risky sexual behavior scores than thei r peers (Langer, Warheit, & McDonald, 2001). Religiosity has additionally been shown to be associated with age of sexual debut (Etzkin, 2004; Jackson, 2005). A study by Mancini and Huebner (2004) suggested that time-use protective factors would decrease adolescents risk-taki ng behaviors, including engaging in sexual intercourse. The authors specified time-use protectiv e factors for this study as str uctured activities that involve
41 formal organizations in order to be successfully completed (for example, extracurricular sports activities and clubs) (p. 650). Results showed that greater invol vement in structured time-use was related to less overall risk behavior. In a ddition, greater involvement in structured time-use was associated with spending more time with friends, having one good frie nd, having an adult to talk to, being more attached to school, and havi ng better grades, among other variables. Mancini and Huebner (2004) conclude that the strong asso ciation between participation in structured activities and almost all variables included in this study atte st to its relevance in understanding youth development and youth risks. They go on to say that, In a sense participation in structured activities provides a context through which positive youth development goals might be met, such as learning cooperative decision-mak ing, interacting with diverse youth and adults, forming strong relationships, and connec ting with caring adults (p. 662). The conclusions reached through this study are applicable to many t ypes of structured activities, including scho ol-related non-sport act ivities such as band, non-school related clubs such as Scouts, volunteer work such as worki ng with the elderly, chur ch or other religiousrelated activities such as yout h group, or school or community -based sports (Mancini & Huebner, 2004). Although research has been conduc ted on each of these main groups, very little has been done to examine the possible effects on adolescent behavior of high-level participation in sports. This particular type of activity is especially important given that nationwide, 57.6% of high school students reported pl aying on one or more sports teams in the past 12 months (Grunbaum, Kann, Kinchen, Ross, Hawkins, Lowry, et al., 2004). The next portion of this review will examine the literature on adolescent participation in sports. Athletic Participation Before reviewing the literature on athletic participation, it is impor tant to understand how many individuals this issue affects and who those individuals are. Alt hough the fact that over
42 half of United States adolescents played on one or more sports teams in the past 12 months (Grunbaum, Kann, Kinchen, Ross, Hawkins, Lowry, et al., 2004) is important, this statistic alone does not provide the full picture. The prevalence of having played on one or more sports teams was higher among male (64%) than female student s (51%). In addition, the prevalence was higher among white (60.8%) than black (53.2% and Hispanic (49.5%) students. More 9th graders (60.3%) than 12th graders (54%) played on one or mo re sports teams (Grunbaum, Kann, Kinchen, Ross, Hawkins, Lowry, et al., 2004). The research on the effects of athletic participation on adolescents is limited and somewhat inconsistent. Some evidence suggests that adolescents who play sports enjoy many positive benefits. Results from a recent st udy by Kirkcaldy, Shephard, and Siefen (2002) demonstrate substantial relationshi ps between regular practice of endurance sport and attitudes, personality, scores for physical and psychologi cal well-being, and the adoption of a healthy lifestyle. A strong link was observed between part icipation in endurance s port and the type of personality that is resistant to addictions. The study also reported an observed link with linguistic competency, which could indicate an underlying high-achie ver personality. In addition, in response to the positive evaluation of a nationa l intervention program that uses sport participation as a ve hicle to enhance life skill development for urban youth, Petitpas, Van Raalte, Cornelius, and Presbrey (2004) s uggest that, Sport participation does provide, however, numerous opportunities for youth to le arn about themselves, to form important relationships with peers and adu lts mentors, and to experience th e benefits of setting goals and working hard to achieve them (p. 333). One interesting factor related to being identified as an athlete in adolescence, one that is neither inherently positive nor n ecessarily negative, is the social perception ascribed to that
43 status. This perception is experienced differe ntly for boys and girls due to traditional gender roles and the behaviors expected of the two gende rs. In an important study of adolescent gender norms and social status among hi gh school students, Suitor and R eavis (1995) examined how the ways in which males and females could gain pr estige in high school changed over time. The data showed that there was relatively little change in gender norms among high school students between the early and late 1980s. For both time periods, significant differences were found between the avenues through which boys and girls c ould gain prestige. In the early 1980s, an important distinction was that for boys, the number one way to acquire prestige was through sports, while this avenue was not suggested as an important one fo r girls. While there was very little change in suggested avenues to gain pres tige across the decade, a particularly interesting development is the increase in girls acquisiti on of prestige through partic ipation in sports and through sexual activity. Other researchers have come to more nega tive conclusions concer ning the effects of athletic participation on adolescent s. Of particular concern is th e increased risk-taking behavior evidenced by adolescent athletes. In their st udy of the health risk behaviors of adolescent participants in organized spor ts, Baumert, Henderson, and Thompson (1998) report that athletes are more likely to wear seatbelts, maintain highe r overall levels of physi cal activity, demonstrate less hopelessness, and exhibit health ier dietary behaviors than thei r peers who are not involved in athletics. However, when compared with nonathle tes, athletes are less li kely to wear a helmet while riding a bicycle or motorcyc le and are more likely to exceed the speed limit while driving. The authors suggest that these resu lts are worrisome, in that the ri sk behaviors that athletes are more likely to engage in could be in terpreted as thrill-seeking behavior.
44 A recent study of adolescent peer crowd aff iliation corroborated the finding that athletes, or those individuals identifying as jocks, report relatively high rates of risk-taking behaviors, suggesting that they may be at risk for nonint entional injuries, even outside of athletic competition (La Greca, Prinstein, & Fetter, 2001). The thrill-seeking and even reckless behavior evidenced by athletes suggests a propensity toward risk-taking and sensati on seeking. This could be a dangerous association. Re search on adolescents suggests th at high sensation seeking is related to a variety of risk behaviors, including gambling (Gupta, Derevensky, & Ellenbogen, 2006), cigarette smoking (Frankenberger, 2004 ), nonmedical prescr iption stimulant use (Herman-Stahl, Krebs, Kroutil, & Heller, 2006), and heavy drinking (vanBeurden, Zask, Brooks, & Dight, 2005), as well as sexual activity, as was discussed earlier in this chapter. Athletic Participation and Adolescent Sexual Behavior To date, very little research has been c onducted to explore the possible relationship between athletic participation and adolescent sexual behavior. What has been done suggests some interesting associations. In a study of adolescent peer crowd affiliation and health-risk behaviors, jocks were not found to engage in high rates of health-risk behaviors (La Greca, Prinstein, & Fetter, 2001). Jocks did, however, report high rates of general risk taking behaviors, have a high proportion of friends who engaged in casual sex, and tend to report more casual sex themselves. The authors conclude that, Athleti cally oriented teens tend to be sexually active (e.g., 59% were sexually active, compared with 42% of the teens overall) and may be engaging in risky sexual behaviors; th ese teens also view themselves as popular and may have more opportunities to find sexual partners than other teens (p. 140). In one of the first studies to examine the po ssible link between adoles cent sexual behavior and athletic participation, Smith and Cald well (1994) asked high school student sports participants and nonparticipants whether they had ever had sexual intercourse. They found that
45 overall, 60.6% of high school spor ts participants reported that they had experienced sexual intercourse, compared to only 41.8% of nonparticip ants who had ever had sexual intercourse. The authors are quick to defend yout h athletic participation in spite of these findings, stating that, We are not suggesting that partic ipation in sports is dysfuncti onal. Rather, these findings indicate a need to (a) more fu lly understand the reasons why high sc hool athletes are more likely to be sexually active than thei r contemporaries, and (b) devel op effective interventions to promote sexual responsibility among these youth (p. 72). An important theme apparent within the li terature on this topic is that of gender differences. In an early and extensive examination of sport and sexual behavior, Miller, Sabo, Farrell, Barnes, and Melnick (1998) found that gender was an influe ntial factor: while athletic participation was associated with lower rates of sexual activity for females, the opposite was true for males. The authors propose a conceptual model for these results based on their reasoning that, Athletic participation confirms traditional gender scripts for boys, while disconfirming traditional scripts for girls. Since these scripts promote (if not overtly) the exchange of female sexual favors for male prestige, at hletic participation indirectly increases sexual activity for boys and decreases it for girls. Sport also enhances social status for both genders, giving male and female athletes greater power in negotiating se xual outcomes in their relationships (p. 120). A related study examining gender and racial di fferences in jock identity and sexual risk found that the jock identity was a disproportiona tely male characteristic and more prevalent among white than black adolescents (Miller, Farrell, Barnes, Melnick, & Sabo, 2005). Males who claimed a jock identity reported more freque nt dating than their nonjock peers, while the dating frequency of female jock s did not differ significantly from that of female nonjocks. Athletic activity was significantly associated wi th lower levels of sexual risk for white but not
46 for black adolescents. The authors emphasize that the correlation between athletic participation (objective measure) and jock identity (subjectiv e measure) is fairly weak and, thus, the two constructs are not interchangeable. They should be examined independently because, Whereas time spent playing sports constituted a buffer agai nst sexual risk, identification with the jock label was actually associated with highe r levels of sexual risk-taking (p. 132). Additional studies focus entirely on female at hletes, leading some to the conclusion that athletic participation serves as a protective fact or for adolescent girls when it comes to sexual behavior. In their study compari ng high-performance female athlet es to the general adolescent female population, Savage and Holcomb (1999) found that six risk-taking behaviors were significantly different between th e two groups: (1) ever engaged in sexual intercourse; (2) age of first sexual intercourse; (3) number of lifetime sexual partners; (4) current sexual activity; (5) condom use; and (6) contracting an STD. With the exception of contracting an STD, the U.S. females reported engaging in risk-taking behavi ors more than the high performance athletes. The higher prevalence of STDs among the athletes may be explained by th e fact that athletes have a greater awareness of their bodies and/ or by the under-reporting of STDs in the U.S. Lehman and Koerner (2004) found that am ong adolescent women who had ever engaged in sexual intercourse, those who were more involve d in sports during high school were less likely to have engaged in sexual risk -taking behaviors and more likel y to have experienced greater sexual/reproductive health. The authors also suggested possible mediator variables. Self empowerment/self efficacy appeared to be a medi ator in the following associations: (a) sport involvement and sexual risk-taking behavior; (b) sport involvement and sexual/reproductive health-seeking behavior; and (c) sport involveme nt and sexual/reproductive health. Functional
47 body orientation emerged as a mediator in the a ssociations between (a) sport involvement and sexual health-seeking behavior, and (b) sport involvement and sexual/reproductive health. Not all research on sport and adolescent sex has demonstrated a link. In their study of possible causal structures in the relationship between participation in athletics and female sexual risk behavior, Dodge and Jaccard (2002) concluded that athletics does not have any direct effect on teenage pregnancy. Although results showed a relationship between participation in sports and the occurrence of a pregnancy in the following 12 months, statis tical analyses suggested that the effect might be spurious. Specifically, female s who participate in athletics are likely to be younger, from highly educated families, and white. These characteristics represent a demographic profile of girls who are also less likely to become pregnant, regardless of athletic participation. In this study, the best explana tion of the association be tween sport participation and sexual activity was a mediational model in whic h pregnancy attitude is the primary mediator of the effect of sport participation. According to the authors, It would seem that females who participate in sports develop, on average, more negative attitudes towa rd the prospects of becoming pregnant than nonathletes a nd that these attitudes in turn make it less likely that they will have sexual intercourse (p. 62). Further complicating the relationship betw een adolescent sexual activity and sport participation is the possible infl uence of various sport-related f actors that may affect the overall experience of being an athlete. Such factors mi ght cause one athlete to experience very different effects, whether positive or negative, stemming from his sport participation than an athlete in a different sport context. For example, type of sport has been linked to body image and dieting problems (Crissy & Honea, 2006), as well as pr oblem alcohol consumption (Martens, Watson, & Beck, 2006), with participants in different sports reporting varying levels of these behaviors. To
48 further explore the impact of type of sport, it is helpful to examine the identifying characteristics typifying specific sports. Two such characterist ics are level of contact and team vs. individual sport. Though limited, some research has been c onducted to look at the e ffects of these factors on athletes. Contact sport athl etes, both male and female, have been shown to demonstrate greater physical aggression outside of sport than athletes in n oncontact sports (Nixon, 1997). In a more recent study, Rudd and Stoll (2004) sought to examine the type of character possessed by college students. They found that team sport pa rticipants reported higher levels of social character (qualities such as team work, loyalty, self-sacrifice, work ethic, and perseverance) than either individual sport partic ipants or nonathletes. In addition to these sport characteristics, factor s related to athlete stat us might play a role in the sport participation experien ce. Athletic identity has been shown to be a mediating factor in the relationship between sport participation an d adolescent adjustment in multiple studies. Guest and Schneider (2003) found that the value ga ined through involvement in sport is affected by the context in which an indivi dual athlete experiences his/her s port participation; context, in this case, consisted of school, community, and at hletic identity. More over, Ryska (2002) found evidence supporting the importance of athletic identity to student athletes overall perceptions of competence (scholastic, social, vocational, and beha vioral). Lastly, the in tensity with which an adolescent participates in hi s/her chosen sport has been li nked to adolescent well-being. Specifically, Pyle, McQuivey, Brasington, and Steiner (2003) found that high school student athletes who participated in s port at an organized, intensive level experienced fewer mental health problems, fewer total risks, and fewe r eating and dietary problems than high school students who participated in sport at a lower intensity. Though adoles cent sexual behavior has not yet been examined in relation to any of these sport or athlet e status factors, it is possible that
49 differences in sport experience related to thos e factors may have an impact on the relationship between athletic partic ipation and sexual debut. Although the research conducted th us far is fascinating and s uggestive, it is clearly far from conclusive. Numerous aspects of the possi ble association between athletic participation and adolescent sexual behavior lack clarity and necessita te further illumination. Summary The initiation of sexual intercourse in adoles cence, although a very personal decision and act, is influenced by a multitude of factors. These factors exist throughou t the various areas of an individuals life, serving as a reminder of th e complexity of adolescent sexual behavior. As Small and Lusters (1994) ecolo gical, risk-factor model of adolescent sexuality asserts, adolescents face risks at all leve ls of their social ecology, which consists of the individual, familial, extra-familial, and macrosystem levels. Further, various factors within these levels can accumulate and interact to lessen or enhance the imp act of the risk factors. This thesis aims to add to the body of literature on sexual debut by examining one such possible influence. This research aims, by looking at adolescent sexu al activity from an ecological perspective, to examine athletic participati on as a possible influence on the d ecision to initiate intercourse. Another aim of this study is to strengthen and expand the body of research on Small and Lusters (1994) ecological approach to a dolescent sexual behavior by expl oring the link between athletic participation and adolescent sexual debut.
50 CHAPTER 4 RESULTS The primary purpose of this study was to examin e the effects of athl etic participation on adolescent sexual debut. This study examin ed various high school experiences, including important factors in the human ecology of adolescents, particular ly athletic participation, as reported by current college student s. Age of sexual debut, along with aspects of related sexual behaviors, was examined in relation to these factors. Multiple statistical tests were conducted to analyz e the data in order to answer the research questions. First, data analysis began with desc riptive analyses to measure the demographics of the sample. Next, bivariate analyses (independ ent-samples t-tests and one-way analyses of variance) were completed to look for differences in both age of sexual de but and sexual activity score by the demographic vari ables and by the various inde pendent variables (athletic participation, ecological variables, and sensati on seeking). Finally, multiple regression analyses were conducted to capture the complex nature of the many factors infl uencing sexual decision making. The combination of all of these statistica l tests allowed the resear cher to fully consider each of the research questions under ex amination in this research study. Descriptive Results In order to provide a comprehe nsive view of the study sample, this chapter begins with the demographics and other descriptiv e results. This discussion provide s a more holistic view of the participants, which is helpful in examining the biva riate and multivariate analyses that follow. Sample Demographics Participants in the study sample ranged in age from 18.00 to 25.83 years, with a mean age of 20.46 years (SD 1.54). The median age was 20.08 years. There were a total of 121 male
51 participants (29.3%) and 292 female participants (70.7%) in this study. As compared to the University population as a whole, this sample includes a larger proportion of female students. Table 4-1 Gender frequency Study sample University undergraduates Gender Frequency Valid percent Frequency Valid percent Male 12129.316,678 46.1 Female 29270.719,484 53.9 Total 413100.036,163 100.0 The majority of participants (n = 234) included in the study sample reported their race/ethnic origin as White ( 57.4%). Black respondents made up 16.9% of the study sample (n = 69. A total of 58 respondents (14.2%) identified their race/ethnic orig in as Hispanic. Thirty-four respondents reported their race/et hnic origin as Asian/Pacific Is lander (8.3%). The remaining participants (n = 13, 3.2%) indicated that they iden tified with some other r ace/ethnic origin than the given answer choices. For those who wrote their race/ethnic origin in the space provided, examples of responses include West Indian, South Asian Indian, Indi an, Middle Eastern, and biracial. This racial profile is consistent with that of the University undergraduate population, though the study sample includes a slightly higher percentage of Black students and slightly lower percentage of White students. Table 4-2 Race/ethnic origin frequency Study sample University undergraduates Race Frequency Valid percent Frequency Valid percent Hispanic 5814.24,693 13.0 Asian/Pacific Islander 348.32,625 7.2 White (non-Hispanic) 23457.423,809 65.8 Black (non-Hispanic) 6916.93,456 9.6 Other 133.21,580 4.4 Total 408100.036,163 100.0 Because of the low number of respondents indi cating Asian/Pacific Islander or Other as their race/ethnic origin, a new race/ethnic origin variable was created by collapsing these two categories into a new Other group. The resul ting groups were: White (56.1%), Black, (16.5%),
52 Hispanic (13.9%), and Other (11.3 %). All analyses that include race/ethnic origin group were run using this new variable. The great majority of the part icipants (n = 307) included in the study sample reported their religious affiliation as Christia n (74.9%). The second most chosen response (n = 47) was No Affiliation (11.5%). A total of 31 respondents indi cated their religious affiliation as Jewish (7.6%), while a small portion of the study sample (2.0%) reported their religious affiliation as Muslim (n = 8). The remaining participants (n = 17, 4.1%) indicated that they identified with some other religious affiliation not among the answer choices. For those who wrote their religious affiliation in the space provided, ex amples of responses include Buddhism, Hinduism, Eastern Orthodox, Bahai, and Wiccan. Table 4-3 Religious affiliation frequency Religious affiliation Frequency Valid percent Jewish 317.6 Muslim 82.0 Christian 30774.9 Other 174.1 No affiliation 4711.5 Total 410100.0 Because of the low number of respondents indicating Jewish, Muslim, Other, or No Affiliation as their religious affiliation, a ne w religious affiliation variable was created by collapsing these four categories into a new Ot her group. The resulting groups were: Christian (73.6%) and Other (24.7%). All analyses that in clude religious affiliation group were run using this new variable. Recognizing the wide variation that exists within the Chris tian faith, a space was provided for respondents identifying their religious affiliation as Christ ian to indicate the specific denomination to which they belong. Twenty-two different denominations were identified through this process. Because many of the denomination categories contained only a few
53 respondents, they were grouped into more manag eable categories as fo llows: Baptist; Catholic; Lutheran/Presbyterian/Methodist; a nd Non-denominational/Other. Table 4-4 Christian denomination frequency Denomination Frequency Valid percent Baptist 4717.8 Catholic 9536.0 Lutheran/Presbyterian/Methodist 4015.2 Non-denominational/other 8231.1 Total 264100.0 Athlete Status Respondents were asked to indi cate their status as an athl ete from the following four choices: I was an athlete in high school and I am now a collegiate athlete; I was an athlete in high school and I now participate in club or intramural sports; I was an athlete in high school and I am not currently active in sports; and I was not an athlete in high sc hool. A total of 123 participants (31.1%) reported that they were not an athlete in high school, while 152 participants reported that they were an athlete in high school but were not currently active in sport (38.4%). The remaining respondents were athletes in high school a nd continue to be active in sport in college, either at the collegiate level (n = 12, 3.0%) or at the club/intramural level (n = 109, 27.5%). Table 4-4 Athlete status frequency Athlete status Frequency Valid percent I was an athlete in high school and I am now a collegiate athlete competing for the University. 123.0 I was an athlete in high school and I now participate in University club or intramural sports. 10927.5 I was an athlete in high school and I am not currently active in sports. 15238.4 I was not an athlete in high school. 12331.1 Total 396100.0 For analysis purposes, two new athlete status variables were created. One variable, designed to capture the distinct ion between respondents who have remained active in sport and those who have not, consists of three groups: Cu rrent Athlete (respondents who were athletes in
54 high school and are still at hletes, either at the co llegiate or club/intramural level 29.0%), High School Athlete (respondents who were athletes in high school and are no longer active in sport 36.5%), and Nonathlete (respondents who were not athletes in high school 31.2%). Another variable was created to look simply at athletes vs. nonathletes. This variable consists of two groups: Athlete (respondents who were athletes in high school, regardless of current athletic participation 70.3%) and Nonathlete (respondents who were not athletes in high school 29.7%). Unless it is clearly stated that the 2-group athl ete status variable wa s used, the analyses described in this chapter we re conducted using the 3-group at hlete status variable. Sexual Debut A total of 252 respondents (64.3%) reported already having their sexual debut. The remaining 35.7% of the sample (n = 140) indicated th at they had not yet had their sexual debut. Table 4-5 Sexual debut frequency Sexual debut Frequency Valid percent No 140 35.7 Yes 252 64.3 Total 392 100.0 Although sexual debut is the focus of this st udy, it is impossible to understand the sexual decision making of adolescents wi thout also considering other re lated sexual behaviors. Thus, respondents were also asked about other sexual behaviors, includi ng the age at which they first engaged in several behaviors on a sexual activity continuum. The mean and median ages at first time, as well as the range in age of first time for each of the eight sexual behaviors appear in Table 4-6 below. Participants were also asked to indicate from 0 (Never) to 5 (A Lot) how often they had engaged in each of the behaviors. For kissing, 4 .2% (n = 17) said they had never engaged in the behavior, while 52.2% (n = 210) said they had enga ged in the behavior a lot. For French kissing,
55 6.7% (n = 27) said they had never engaged in th e behavior, while 44.5% (n = 179) said they had engaged in the behavior a lot. For touching a partners breast or having your breast touched by a partner, 13.2% (n = 52) said they had never engaged in the behavior, wh ile 26.1% (n = 103) said they had engaged in the behavior a lot. For touching a partners penis or having your penis touched by a partner, 19.4% (n = 76) said they had never e ngaged in the behavior, while 26 .1% (n = 103) said they had engaged in the behavior a lot. For touching a partners vagina or having your vagina touched by a partner, 20.4% (n = 80) said they had never e ngaged in the behavior, while 23.4% (n = 92) said they had engaged in the behavior a lot. For performing oral sex, 33.3% (n = 131) said they had never enga ged in the behavior, while 16.5% (n = 65) said they had engaged in th e behavior a lot. For receiving oral sex, 27.7% (n = 109) said they had never engaged in the behavior, while 21.4% (n = 84) said they had engaged in the behavior a lot. For sexual in tercourse, 35.7% (n = 140) said they had never engaged in the behavior, while 25.5% (n = 100) said they had e ngaged in the behavior a lot. Table 4-6 Debut age for sexual behaviors Sexual behavior Mean age at first time (years) Median age at first time (years) Range in age at first time (years) Kissing 14.0514.00 5.00 21.50 French kissing 14.8314.75 6.00 21.75 Touching a partners breast or having your breast touched by a partner 15.8716.00 11.00 21.08 Touching a partners penis or having your penis touched by a partner 16.3316.00 12.00 21.00 Touching a partners vagina or having your vagina touched by a partner 16.4516.00 12.00 -22.00 Performing oral sex 16.8317.00 12.00 24.00 Receiving oral sex 16.8817.00 12.00 24.00 Sexual intercourse 17.1917.00 13.00 24.00
56 Age of Sexual Debut The mean age of sexual debut for the study sample was 17.19 years and the median age was 17.00 years. Sexual debut ages ranged from 13.00 to 24.00 years. Bivariate Analyses Bivariate analyses were conduc ted to determine the effect of demographics (gender, religious affiliation, race/ethnic origin) on the independent variables (a thletic participation, ecological variables, sensation seeking). Bivari ate analyses were also conducted to determine the relationship between sexual be havior dependent variables (age of sexual debut and score on the sexual activity scale) and demographic variab les, as well as the relationship between the dependent variables and several i ndependent variables. Significan t results are reported in this section. Tables of the significant results can be found in Appendix D. Demographics One aim of this study was to determine the effects of numerous variables (athletic participation, ecological variable s, sensation seeking) on sexual behavior. If any conclusions regarding the effects of these va riables are to be made, it is important to first understand the possible impact demographics have on these fact ors. Thus, analyses were conducted to examine each of the independent variables by each of the demographic variab les. The significant findings are reported in this section. In this study, gender was the most influential demographic characterist ic. Several of the ecological variables differed by gender. Fe male respondents scored higher than male respondents on the Risk Avoidance (t = -3.452, p = .001), Parental Moni toring (t = -4.325, p = .000), and Values (t = -2.667, p = .008) indices. Fe males also agreed more with the statement My schoolwork was very important to me (t = -5.569, p = .000). Males, however, reported higher scores on the Popularity index (t = 5.336, p = .000). Additionally, scores on the Brief
57 Sensation Seeking Scale (BSSS) differed by gender (t = -4.1 21, p = .000), with males reporting lower scores (indicating higher propensity for sensation seeking). Gender also had an impact on the sport partic ipation variables. Females received higher scores (indicating less agreement and, thus, less identification with athlete status) on the total Athletic Identity Measurement Scale (AIMS) scale (t = -2.442, p = .015), as well as on the Exclusivity (t = -2.529, p = .012) and Negative Affectivity (t = -2.044, p = .042) domain scales. Males reported a much higher intensity of spor t participation, as measured by hours per week spent participating in sport (t = 3.209, p = .001) Males reported an average of 17.12 hours per week in sport, while females reported an aver age of 13.97 hours per week. Respondents choice of primary sport differed by gender as well (Pearson Chi-Square = 47.319, p = .000). Lastly, level of contact differed by gender, as show n by Cross-tabs reported below (Table 4-7). Table 4-7 Cross-tabulation of level of contact and gender Level of contact Gender Noncontact Contact Collision Total Male Frequency 393620 95 % of Gender 41.137.921.1 100.0 Female Frequency 143490 192 % of Gender 74.525.50.0 100.0 Total Frequency 1828520 287 Pearson Chi-Square = 59.571, p = .000 Religious affiliation was significantly related to several independent variables. Christian respondents scored higher than Ot her religious affiliation responde nts on the Risk Avoidance (t = 3.328, p = .001), Connection to School and Othe rs (t = 5.571, p = .000), and Values (t = 12.342, p = .000) indices. BSSS score also differed by re ligious affiliation (t = 1.980, p = .048), with Christian respondents receiving a higher mean sc ore (indicating lower propensity for sensation seeking) than those reporting their religious affiliation as Other.
58 Race/ethnic origin seemed to have an effect on some of the independent variables. The race groups differed in mean score on the Connec tion to School and Others (F = 7.588, p = .000) index, with the Other race categor y scoring lower than all of th e other categories. BSSS score also differed by race (F = 5.595, p = .001), with Bl ack respondents receiving the highest mean score, followed by respondents reporting their race as Hispanic, White, and Other. ANOVA used to look for differences in athletic identity by race group revealed significant differences in mean total AIMS (F = 4.433, p = .005), Social Identity (F = 3.759, p = .011), and Negative Affectivity (F = 5.495, p = .001) scores. Post Hoc an alyses revealed that th e main differences for each of these three variables existed between White respondents and Black respondents. Sexual Behavior Age of Sexual Debut To explore possible relationships between ag e of sexual debut and demographics, t-tests and ANOVA were conducted. No si gnificant differences were found. To determine the extent to which personal experiences affected age of sexual debut, ANOVA was conducted for the BSSS and for each of the six ecological variables indices, as well as for the two remaining individual ecological variables items. For those variables that are measured by a score, simply entering the variab le as it was would have resulted in too many groups for the ANOVA test to handle. Thus, ne w variables were created for BSSS and each of the ecological indices by groupi ng scores into four groups based on quartiles. These new variables were only used in the ANOVA for age of sexual debut and sexual activity score. A significant difference in age of sexual debut wa s found by score on both the Risk Avoidance (F = 8.984, p = .000) and Values (F = 5.862, p = .001) indices with those reporti ng lower scores also reporting younger ages of sexual debut. Lastly, a significant difference in age of sexual debut was found for the item I felt like I received a goo d education in high school (F = 3.783, p =
59 .005). Post Hoc analyses revealed that th e differences existed between respondents who answered Neutral in response to this stat ement (mean age of sexual debut 16.53 years) and those who answered Strongly Agree (mean age of sexual debut 17.44 years). To better understand what effect sport partic ipation might have on ag e of sexual debut, ttests and ANOVA were completed to look for relati onships between this de pendent variable and sport participation variables (athle tic identity, intensity of sport pa rticipation, type of sport, level of contact, and team vs. individual sport). No significant differences were found. Because the literature review suggested that different relationships between sport participation and age of sexual debut existed for males and females, additional t-tests and ANOVA were conducted to examine age of debut by athlete status, using both the 3-group and the 2-group variables, for males separately and for females separately. No significant differences were found. The researcher then selected only the athletes and conducted a t-test to explore age of sexual debut by gender. Again, no significant differences were found. Sexual Activity Though age of sexual debut is an important ma rker in understanding sexual behavior, it is equally important to examine total sexual activity. Possible effects of demographics on sexual activity were examined using t-tests and ANOV A. A significant difference in mean sexual activity score was found by gender (t = 2.683, p = .008), with males reporting a mean score of 27.70 and females reporting a mean score of 24.78. Additional t-tests and ANOVA were conducted to examine possible relationships between sexual activity and various personal experiences including sensation seeking and ecological variables. ANOVA conducted for the BSSS and ec ological variables indice s used the variables with scores grouped into quartiles, as described previously for age of sexual debut. A significant difference in mean sexual activity score wa s found by BSSS score (F = 8.064, p = .000), with
60 those scoring lower on the BSSS (indicating higher propensity for sensation seeking) reporting higher sexual activity scores. A significant diffe rence in mean sexual activity score was found by score on the Risk Avoidance (F = 30.680, p = .000), Parental Monitoring (F = 11.867, p = .000), and Values (F = 12.305, p = .000) indices. Fo r each of these variables, those reporting higher scores on the indices report ed lower sexual activity scores. To examine to what extent sport participat ion might influence overall sexual activity, ttests and ANOVA were conducted for athletic identit y, intensity of sport participation, type of sport, level of contact, and team vs. individual sp ort. No significant differences were found. To further explore the possible intera ction of sport participation, ge nder, and sexual activity, several more t-tests and ANOVA were completed. First, separate analyses were conducted for males only and females only to look for differences in sexual activity score between athletes and nonathletes. No significant di fferences were found. The researcher then selected only the athletes and conducted a t-test examining sexual activity by gender. A significant difference was found (t = 2.936, p = .004), with male athletes re porting higher sexual activity (mean score 28.48) than female athletes (mean score 24.96). This pattern mirrors that of the whole sample. Athlete Status One of the primary goals of this study is to examine differences between athletes and nonathletes. To begin this explor ation, Cross-tabs of athlete stat us and each of the demographic variables were conducted (Table s 4-8, 4-9, and 4-10 below). Si gnificant relationships were found between athlete status and gender, athlete stat us and religious affilia tion, and athlete status and race/ethnic origin group. To determine whether ecological variab les differed by athl ete status, ANOVA was conducted to look for differences in mean score on each of the six indices, as well as differences in response to the two remaini ng items. Significant differences were found in mean Connection
61 to School and Others score (F = 22.016, p = .000), with current athletes reporting the highest mean score, followed by high school athl etes, and, lastly, nonathletes. Table 4-8 Cross-tabulation of gender and athlete status Athlete status Gender Current High school Nonathlete Total Male Frequency 583324 115 % of Gender 50.428.720.9 100.0 Female Frequency 62117100 279 % of Gender 22.241.935.8 100.0 Total Frequency 120150124 394 Pearson Chi-Square = 30.832, p = .000 Table 4-9 Cross-tabulation of religious affiliation and athlete status Athlete status Religious affiliation Current High school Nonathlete Total Christian Frequency 9611983 298 % of Religion 32.239.927.9 100.0 Other Frequency 243040 94 % of Religion 25.531.942.6 100.0 Total Frequency 120149123 392 Pearson Chi-Square = 7.173, p = .028 Table 4-10 Cross-tabulation of race/e thnic origin and athlete status Athlete status Race/ethnic origin Current High school Nonathlete Total White Frequency 849150 225 % of Race 37.340.422.2 100.0 Black Frequency 112926 66 % of Race 16.743.939.4 100.0 Hispanic Frequency 112124 56 % of Race 19.637.542.9 100.0 Other Frequency 121020 42 % of Race 28.623.847.6 100.0 Total Frequency 118151120 389 Pearson Chi-Square = 26.350, p = .000 In examining the possible relationship between athlete status and sensation seeking, an ANOVA could not be used for the 3-group athlete status variable because it did not meet the assumption of homogeneity of variances (Leven e statistic = 4.010, p = .019). Thus, a t-test was used to examine differences in BSSS score fo r athletes and nonathletes using the 2-group
62 variable. This analysis rev ealed a significant difference (t = -3.188, p = .002), with athletes reporting lower scores (indicati ng higher propensity for sensati on seeking) than nonathletes. To further the exploration of the importance of athlete status, ANOVA was conducted to reveal possible differences in athletic identity between athletes who are currently involved in sport and those whose sport partic ipation ended in high school. Si gnificant differences in mean Exclusivity and Negative Affec tivity scores were found betw een the two groups. An ANOVA could not be used to examine relationships between total AIMS score and Social Identity score and athlete status because these tests did not m eet the assumption of homogeneity of variances (Levene statistic = 4.582, p = .033; Levene stat istic = 14.140, p = .000, respectively). To address this problem, a new variable was created by gr ouping AIMS scores into four groups based on quartiles. This was also completed for the Social Identity variable. ANOVA was completed using these new variables. In th is manner, significant differences in mean total AIMS and Social Identity scores between current athletes and high school athletes were revealed (mean scores and F-values reported in table 4-11 below). Table 4-11 ANOVA for AIMS by athlete status Social identity (mean score) Exclusivity (mean score) Negative affect (mean score) Total AIMS (mean score) Athlete status (F) 37.439*** 21.165*** 18.095*** 33.295*** Current athlete 6.468.335.60 20.39 H.S. athlete 9.3410.057.18 26.57 *** Significant at .001 level Multivariate Analyses Although the bivariate analyses revealed many interesting relationships between the sexual behavior dependent variable s (age of sexual debut and ove rall sexual activity) and both the demographic and the various in dependent variables, the results present such relationships as if they occur in a vacuum. This does not, how ever, reflect the way decisions regarding sexual
63 behavior are actually made. Therefore, multivar iate regression was used to determine what variables most significantly affect sexual beha vior in a context more reflective of actual conditions and relationships. Sepa rate regressions were run for age of sexual debut and overall sexual activity. Age of Sexual Debut The dependent variable for this multiple regr ession was age of sexual debut (Table 4-12, at the end of this chapter). This regression consisted of four distin ct models (1) sociodemograhic variables only, (2) ecological variables only, (3) sensation seeking items onl y, and (4) all of these variables together. A fifth re duced model was then developed to determine which of these independent variables were most significantly re lated to age of sexual debut. This reduced model allowed for the analysis of the impact of only those factor s that were most important to age of sexual debut, without th e background noise created by the inclusion of nonsignificant variables. Only those responde nts who indicated that they ha d already had their sexual debut were included in this analysis (n = 228). Model 1. The first model included the demographic variables (gender, race/ethnic origin, religious affiliation, and athlete status). Current age was not in cluded in this analysis, as it represented a large num ber of missing cases and was thought by the researcher to confound the age of sexual debut dependent variable. Demogra phic variables with nomi nal data were recoded to create dummy code variables. Race group variables included White Black, Hispanic, and Other Race; the White group was held constant and served as a refe rence group. Religious affiliation groups included Christian and Other Religion, with Other Religion serving as the reference group. The researcher wanted to ensure that breaki ng up the religious affiliation categories into smaller, more di verse groups would not affect the results. Thus, new variables were created for use in the multivariate analyses which included Jewish, Other Religion, No
64 Affiliation, Baptist, Catholic, Lutheran/Methodi st, and Other Christian. Including these variables into the regressions did not change the models. At that point, it was decided to include only the Christian and Other Religion variables to reduce the total num ber of variables under consideration. Lastly, athlete st atus groups included Current Athlete, High School Athlete, and Nonathlete, with Nonathlete se rving as the reference group. The model was not found to be significant and accounted for only about 1% of the variance in age of sexual debut (Adj. R2 = .014). None of the demographic variables we re found to be statistically significant. Model 2. The second model included each of the six ecological variables indices (Risk Avoidance, Parental Closeness, Parental Monitoring, Connection to School and Others, Values, and Popularity), as well as the remaining two i ndividual ecological variables items (I felt like I received a good education in high school; and My schoolwork was very important to me). Model 2 was significant (F = 6.093, p = .000) an d accounted for approximately 15% of the variance in age of sexual debut (Adj. R2 = .154). The following variables were found to be statistically significant: Risk A voidance, Values, and I felt like I received a good education in high school. All three of these variables were positively related to age of sexual debut. As scores on the Risk Avoidance and Values indice s increased, age of sexua l debut tended to be higher. Likewise, as agreement with the statem ent I felt like I received a good education in high school increased, so did age of sexual debut. Model 3. The third model included the four content domains making up the BSSS. For purposes of all multiple regressions reported in th is chapter, BSSS items were reverse-coded so that higher scores corresponded to higher prope nsity for sensation seeking. Model 3 was significant (F = 3.931, p = .004) and accounted for al most 5% of the variance in age of sexual debut (Adj. R2 = .047). The only domain score found to be statistically significant was that for
65 Disinhibition This item was negatively related to ag e of sexual debut; as score on this scale increased, age of sexual debut tended to be younger. Model 4. The fourth model included all of the va riables discussed in the previous three models (demographic variables, ecological variables, sensation seeking variables) together. Model 4 was significant (F = 2.882, p = .000) and acc ounted for over 13% of the variance in age of sexual debut (Adj. R2 = .136). The following variables were found to be statistically significant: Risk Avoidance, Values, and the eco logical statement I felt like I received a good education in high school. As in the earlier mode ls, each of these variables was positively related to age of sexual debut, meaning that a higher scor e was reflected in a late r age of sexual debut. Reduced Model. While Model 4 revealed important results, it is somewhat cluttered by the inclusion of many nonsignifi cant variables. Therefore, a reduced model containing only the significant variables was devel oped. This was accomplished by systematically removing the least significant variable until a parsimonious and highly predictive model was achieved. This final reduced model was significant (F = 13.603, p = .000) and accounted for approximately 14% of the variance in age of sexual debut (Adj. R2 = .142). Three variables were found to be significant. Risk Avoidance, Values, and the ecological statement I fe lt like I received a good education in high school were positively related to age of sexual debut ; the higher the score on these items, the later the age of sexual debut. This initial multiple regression model provid ed an opportunity to closely examine which factors had the most impact on age of sexual debut and to more clearly elucidate the relationships among these various factors. Although this anal ysis was invaluable in better understanding sexual debut, it represents a si ngle act (though a highly important one) at one moment in time, without taking into account the myriad othe r activities leading up to and surrounding it.
66 Additionally, only those respondent s who had already had their se xual debut could be included, thus excluding a significant number of people, a nd their sexual experiences. Therefore, an additional multiple regression analysis was conducted to further examine the sexual behaviors, and those factors most influentia l to those behaviors, of all re spondents in the sample, regardless of level of sexual activity. Sexual Activity The dependent variable for this multivariate regression was score on the sexual activity scale (Table 4-13, at the end of th is chapter). This regression consisted of the same four models (1) sociodemograhic variables only, (2) ecological variables onl y, (3) sensation seeking items only, and (4) all of these variables together; an additional fifth reduced model was also developed to determine which of these factors were most significantly rela ted to sexual activity. The entire study sample was included in this analysis (n = 382). Model 1. The first model included the same dem ographic variables, with the addition of age. Dummy codes for the demographic variables of race, religious affilia tion, and athlete status were included in the same ma nner as detailed above for the ag e of sexual debut regression. Model 1 was significant (F = 2.185, p = .028) and accounted for about 2% of the variance in sexual activity score (Adj. R2 = .024). The only variable that was found to be statistically significant was gender. Males we re likely to have a higher sexua l activity score than females on the sexual activity scale. Model 2. The second model included each of the si x ecological variables indices, as well as the remaining two individual ecological variab les items. Model 2 was significant (F = 16.375, p = .000) and accounted for almost 24% of the variance in sexual activity score (Adj. R2 = .239). The following variables were found to be statistically significan t: Risk Avoidance, Parental Monitoring, Connection to School and Others, and Values. Parental Monitoring, Risk
67 Avoidance, and Values were nega tively related to sexual activity; as scores on these indices went up, the sexual activity score went down. C onnection to School and Others, however, was positively related to sexual activity; as scores on this index went up, so did sexual activity score. Model 3. The third model included the four content domains making up the BSSS. Model 3 was significant (F = 20.564, p = .000) and accounted for approximately 16% of the variance in sexual activity score (Adj. R2 = .163). Two variables were found to be statistically significant. Boredom Susceptibility was nega tively related to sexual activity; as scores on this scale increased, sexual activity score decreased. Disinhib ition was positively related to sexual activity; as scores on this scale increased, se xual activity score also increased. Model 4. The fourth model provided a compre hensive analysis including all of the variables discussed in the previ ous three models (demographic va riables, ecological variables, sensation seeking variables) together. Model 4 was signi ficant (F = 8.218, p = .000) and accounted for approximately 27% of the va riance in sexual activity score (Adj. R2 = .274). The following variables were found to be statisti cally significant: Risk Avoidance, Parental Monitoring, Connection to School and Others, Values, and the sensation seeking domains Boredom Susceptibility and Disinhi bition. As in the earlier mode ls, Risk Avoidance, Parental Monitoring, and Values, as well as Boredom Sus ceptibility, were positiv ely related to sexual activity, while Connection to School and Others and Disinhibition were negatively related to sexual activity. Reduced Model. To more easily identity which factors were most important to overall sexual activity, a reduced model containing only th e significant variables was developed in the same manner as the age of sexua l debut regression. This final reduced model was significant (F = 22.317, p = .000) and accounted for just over 28% of the variance in sexual activity score (Adj.
68 R2 = .281). Six variables were found to be statis tically significant. C onnection to School and Others and the sensation seeking domain Disinhibi tion were positively related to sexual activity; the higher the score on these item s, the higher the sexual activit y score. Risk Avoidance, Parental Monitoring, and Values we re negatively related to sexua l activity; the higher the score on these items, the lower the sexual activity scor e. Additionally, current athlete status was negatively related to sexual activ ity; current athletes reported lower sexual activity scores than nonathletes. In considering the results of these multiple regressions, the researcher was concerned that the high proportion of females in the sample mi ght confound the analysis. Therefore, separate reduced models were examined for males and fe males for both age of sexual debut and sexual activity to determine if they differed in any significant way from each other or from the study sample as a whole. The findings from the fema le-only model basically mirrored those of the full sample model, suggesting no important differences. This also seemed to be true for males, despite the limited statistical power of the male -only models due to the low number of male cases. For both males and females, very simila r patterns corresponding to th at of the full sample model were revealed, supporting the use of the original model. It was determined through these analyses that gender was not a pr oblem and that the original model could serve as a sufficient representation of all individua ls in the full study sample. Athlete Status Because Current athlete status was statistica lly significant in the reduced model, with current athletes reporting generally lower sexual activity scores than nonath letes, the researcher conducted additional multiple regression analyses to further explore the unique influences on the sexual behaviors of athletes and nonathletes (Table 4-14, at the end of this chapter). Four separate reduced models were completed. Th e first model (Sample) is the reduced sexual
69 activity model from the above analysis, conducted w ith the entire sample. This is reported again and included in the table to serve as a comparison for the ot her models. The second model (Nonathletes) is a reduced mode l including all demographic, eco logical, and sensation seeking variables conducted with only the nonathletes. The third model (Athletes) is a reduced model including all demographic, ecological, and sensa tion seeking variables conducted with only the athletes. For these multiple regression analyses, the researcher included all athletes, whether or not they were currently active in sport. Th e fourth model (Athlete s+) is a reduced model including all demographic, ecological, and sensa tion seeking variables, along with the sport participation variables (contact vs. noncontact sport; team vs. i ndividual sport; intensity; AIMS Social Identity; AIMS Exclusivity; AIMS Ne gative Affectivity; and the additional AIMS statement added by the researcher, I was part of the most popular group at school). Nonathletes. The second model, including only nona thletes (n = 112), was significant (F = 9.685, p = .000) and accounted for almost 24% of the variance in sexual activity score (Adj. R2 = .237). Four variables were found to be statistically significant. Risk Avoidance and Parental Monitoring were negatively related to sexual activ ity; as scores on these items increased, sexual activity score decreased. Connection to School and Others and the sensation seeking domain Experience Seeking were positivel y related to sexual activity; as scores on these items increased, so did sexual activity score. Athletes. The third model, including only athlet es (n = 276), was significant (F = 24.915, p = .000) and accounted for approximately 30% of the variance in sexual activity score (Adj. R2 = .302). Five variables were found to be statistically significant Risk Avoidance and Values were negatively related to sexua l activity; those reporting higher scores on these items reported lower sexual activity scores. Connection to Scho ol and Others and sensation seeking domain
70 Disinhibition were positively rela ted to sexual activity; those reporting higher scores on these items also reported higher sexual activity scores Age was also positively related to sexual activity, with older respondents repo rting higher sexual activity scores. Athletes+. The fourth model, again including only athletes (n = 276), added sport participation variables not included in the other multiple regressions. For purposes of this multiple regression, AIMS items were reverse-coded so that higher scores indicated higher levels of agreement and, thus, higher id entification as an athlete. Additionally, the Collision and Contact groups for the level of contact variable were collapsed, and just two groups, Contact and Noncontact, were used for this analysis. Sport participation variables with nominal data were recoded to create dummy code variables. Contact group variables included Contact and Noncontact; the Noncontact group was held constant and served as a reference group. Team vs. individual sport variable s included Team and Individual, with Individual serving as the reference group. The final model was significant (F = 20.947, p = .000) and accounted for approximately 36% of the variance in se xual activity score (Adj. R2 = .366). Eight variables were found to be statistically significant. Age was again positivel y related to sexual activ ity; older respondents reported higher sexual activity scores. The se nsation seeking domain Disinhibition and the AIMS statement I was part of the most popular group at school were positively associated with sexual activity; as scores on these items increased, so did sexual activ ity score. Risk Avoidance, Values, and AIMS Social Identity were negativel y associated with sexual activity; as scores on these times increased, sexual activity score decrea sed. Additionally, contact sport and team sport were negatively associated with sexual activity. Contact sport players reported generally lower sexual activity scores than noncontact sport player s, while team sport participants reported generally lower sexual activity scores than individual sport participants.
71 Summary The various analyses presented in this chapter have helped to explore and bring to light the many relationships between demographic variables, athletic participation, ecological variables, and sensation seeking, and both overall sexual activ ity and age of sexual debut. Particularly intriguing findings were revealed regarding the importance of ge nder, propensity for sensation seeking, and many of the ecological variables. Sexual behaviors, examined through the consideration of both age of sexua l debut and overall sexual activit y, was shown to be influenced by many different factors, with some unexpected relationships. The following chapter will discuss how these results reflect on the resear ch questions, as well as the implications the findings of this study have for practice, policy, and future research.
72 Table 4-12 Comparison of four multivariate models on age of sexual debut Standardized regression coefficients Model 1 Model 2 Model 3 Model 4 Reduced Demographic variables Gender (males = 1, females = 0) .033 -.104 Race/ethnic origin Black -.045 -.097 Hispanic -.035 .025 Other race/ethnic origin -.025 -.025 Religious affiliation Christian .128 -.098 Athlete status Current athlete -.014 -.005 High school athlete .010 .058 Ecological variables Risk avoidance .196** .156* .183* Parental closeness -.026 -.104 Parental monitoring .005 .014 Connection to school and others -.130 -.081 Values .303*** .328*** .250*** Popularity -.056 -.020 Received good education .212** .173* .140* Schoolwork important -.024 .017 Sensation seeking Experience seeking -.073 -.095 Boredom susceptibility .118 .081 Thrill and adventure seeking -.095 -.081 Disinhibition -.187* -.049 R adjusted -.014 .154*** .047** .136*** .142*** Cases 218 224 240 228 228 Significant at .05 level ** Significant at .01 level ***Significant at .001 level
73 Table 4-13 Comparison of four multivariate models on sexual activity Standardized regression coefficients Model 1 Model 2 Model 3 Model 4 Reduced Demographic variables Age -.021 .048 Gender (males = 1, females = 0) -.142* -.011 Race/ethnic origin Black -.059 -.031 Hispanic .017 .007 Other race/ethnic origin -.036 -.007 Religious affiliation Christian -.095 .028 Athlete status Current athlete .001 -.069 -.092* High school athlete .116 .017 Ecological variables Risk avoidance -.324*** -.243*** -.245*** Parental closeness -.037 -.041 Parental monitoring -.122* -.126* -.138** Connection to school and others .172*** .172*** .152*** Values -.188*** -.150** -.118* Popularity .062 .017 Received good education -.059 -.076 Schoolwork important .040 .062 Sensation seeking Experience seeking .088 .069 Boredom susceptibility -.135* -.105* Thrill and adventure seeking -.072 -.085 Disinhibition .447*** .281*** .268*** R adjusted .024* .239*** .163*** .274*** .281*** Cases 384 391 403 382 382 Significant at .05 level ** Significant at .01 level ***Significant at .001 level
74 Table 4-14 Comparison of reduced multiv ariate models on sexual activity Standardized regression coefficients Sample Nonathletes Athletes Athletes+ Demographic variables Age .167*** .157*** Gender (males = 1, females = 0) Race/ethnic origin Black Hispanic Other race/ethnic origin Religious affiliation Christian Athlete status Current athlete -.092* High school athlete Ecological variables Risk avoidance -.245*** -.266* -.290*** -.224*** Parental closeness Parental monitoring -.138** -.260* Connection to school and others .152*** .183* .116* Values -.118* -.204*** -.201*** Popularity Received good education Schoolwork important Sensation seeking Experience seeking .188* Boredom susceptibility Thrill and adventure seeking Disinhibition .268*** .239*** .261*** Athletic participation Contact -.132* Team/individual -.135* Intensity AIMS social identity -.121* AIMS exclusivity AIMS negative affectivity Popular group .265*** R adjusted .281*** .237*** .302*** .366*** Cases 382 112 276 276 Significant at .05 level ** Significant at .01 level ***Significant at .001 level
75 CHAPTER 5 DISCUSSION Adolescence, the tumultuous period between childhood and adulthood, can be a challenging and frightening time even under the best circumstances. It is a time filled with new experiences and newfound freedoms, and with th ese come the opportunity to make some very new, very adult decisions. One of these major d ecisions is the choice of if and when to have sex for the first time. Numerous factors have been shown to influence adolescent sexual debut, including perceived goals (Michels, Kropp, Eyre & Halpern-Felsher, 2005), peer affiliation (Smith and Guthrie, 2005), and involvement in st ructured time-use activities, such as sport (Mancini & Huebner, 2004). The purpose of this study was to explore the e ffects of athletic part icipation on adolescent sexual debut. Sexual behavior, as measured by age of sexual debut and overall sexual activity, was examined in an ecological context to look for relationships with sport participation factors, ecological variables, and sensation seeking dom ains. This chapter presents a review and discussion of the research questi ons, and offers an interpretation of the results of this study as they relate to the theoretical framework. Cont ributions to the literature are then discussed, followed by implications for future resear ch, youth practitioners, and public policy. Research Questions The following research questions were stated in the operational nul l hypothesis form and were analyzed using t-tests, ANOVA, and multiple regression analyses. The first set of research que stions (RQ 1, RQ 1.1, and RQ 1.2) concerned the effect of variables related to athlete status on age of sexual debut. Research Question 1: Does being an at hlete affect timing of sexual debut? Null Hypothesis: Being an athlete does not affect timing of sexual debut.
76 There were no significant diffe rences in age of sexual debu t by athlete status group. In addition, athlete status was not si gnificant in any of the multiple regression models for age of sexual debut. This null hypothesis was not rejecte d. This finding is surp rising, given that prior studies have found significant relationships betw een athlete status a nd sexual behavior (e.g., Miller, Sabo, Farrell, Barnes, & Melnick, 1998; Savage & Holcomb, 1999; Smith & Caldwell, 1994). The contradictory findings of this study may be due to the specific measure of sexual behavior utilized (age of sexual debut). Previous studi es have instead examined measures such as occurrence of sexual debut, pregnancy, contrace ptive use, number of partners, and incidence of STI contraction. Additionally, previous studies exploring sexual debut have sampled primarily high school students; c onclusions reached through this st udy may differ as a result of the sample of college students. Research Question 1.1: Does intensity of sport participation affect tim ing of sexual debut? Null Hypothesis: Intensity of sport participa tion does not affect timing of sexual debut. There were no differences in age of sexual de but by intensity of spor t participation (hours per week spent par ticipating in sport). This null hypothesis was not rejected. Research Question 1.2: Does level of athlet ic identity affect timing of sexual debut? Null Hypothesis: Level of athletic identity does not affect timing of sexual debut. There were no significant diffe rences in age of sexual de but by score on the Athletic Identity Measurement Scale (A IMS) or by score on the three AIMS domain scales (Social Identity, Exclusivity, and Nega tive Affectivity). This null hypothesis was not rejected. The second set of research que stions (RQ 2, RQ 2.1, and RQ 2.2) explored the effect of variables related to sport type on age of sexual debut. Research Question 2: Does type of sport affect timing of sexual debut?
77 Null Hypothesis: Type of sport does not affect timing of sexual debut. There were no differences in age of sexual de but by type of sport. This null hypothesis was not rejected. Research Question 2.1: Does level of c ontact affect timing of sexual debut? Null Hypothesis: Level of contact does not affect timing of sexual debut. There were no differences in age of sexua l debut among respondents participating in Collision, Contact, and Noncontact sports. This null hypothesis was not rejected. Research Question 2.2: Does team vs. indivi dual sport affect timing of sexual debut? Null Hypothesis: Team vs. individual sport does not affect timing of sexual debut. There were no differences in age of sexual debut between participants of team and individual sports. This nu ll hypothesis was not rejected. The third set of research ques tions (RQ 3 and RQ 3.1) explor ed the effect of ecological variables on age of sexual debut. Research Question 3: Do ecological vari ables affect timing of sexual debut? Null Hypothesis: Ecological variables do not affect timing of sexual debut. To answer this question, ANOVA was conducted to look for relationships between age of sexual debut and scores on the six ecological variables indice s (Risk Avoidance, Parental Closeness, Parental Monitoring, Connection to School and Others, Values, and Popularity). Significant differences in sexual debut were found by scores on both the Risk Avoidance and Values indices. For both of these variables, lower scores corresponded with younger ages of sexual debut. Additional ANOVA was conducted to examine possible relationships between age of sexual debut and the two remaining ecologi cal variables items (My schoolwork was very important to me, and I felt like I received a good education in high sc hool). A significant
78 difference was found for the second item, with pa rticipants who indicate d that they strongly agreed with this statement reporting an older age of sexual debut than part icipants who indicated that they felt neutral about this statement. The multiple regression models corroborate the re sults of the bivariate analyses. Model 2, which included each of the ecological variables indices and both of the individual statements, accounted for just over 15% of th e variance in age of sexual debut. Furthermore, all three of the ecological variables significantl y related to age of sexual de but using ANOVA were significant in both Model 2 and Model 4, in which all demogr aphic, ecological, and sensation seeking items were included. Lastly, these three ecological variables remained significant in the reduced regression model for age of sexual debut This null hypothesis was rejected. The specific relationships found in this study between the ecological variables and age of sexual debut were not altogether unexpected, as support for each has been demonstrated through prior research. The finding that personal avoidance of using cigare ttes, alcohol, or marijuana is related to age of sexual debut is in agreement with previous research linking substance use and sexual behavior, including age of sexual debut (e.g, Lowry, Holtzman, Truman, Kann, Collins, & Kolbe, 1994; Tapert, Aarons, Sedlar, & Brown, 2001). The current study found that respondents who felt their parents would have strongly disapp roved of them having sex while in high school reported later ages of sexual debut; this finding is consistent with that of Small and Luster (1994), who reported that negative parental valu es toward sex were a ssociated with a lower chance of experiencing sexual debut in high school. Personal religiosity, demonstrated by statemen ts concerning the importance of religion in an individuals life and the regularity of a ttending religious services, was found to be significantly related to age of sexual debut. Similarly, regular church attendance (Crockett,
79 Bingham, Chopak, & Vicary, 1996) and strong religious values (La nger, Warheit, & McDonald, 2001) have been found to be protective factors ag ainst both initiation of sexual activity and risky sexual behaviors. Furthermore, the results of this study suggest that having positive feelings toward ones education is also a protective f actor, a finding that corr oborates the results of several previous studies. Fo r example, Mancini and Huebner (2004) found that positive school attachment was related to less overall risk-taking behavior, in cluding the initiation of sexual intercourse. Research Question 3.1: Does propensity for sensation seeking aff ect timing of sexual debut? Null Hypothesis: Propensity for sensation seek ing does not affect tim ing of sexual debut. ANOVA was conducted to determine whether a relationship existed between score on the Brief Sensation Seeking Scale (BSSS) and age of sexual debut. No si gnificant results were found. The third multiple regression model for sexual debut, including each of the sensation seeking domains, however, accounted for almost 5% of the variance in age of sexual debut, suggesting that some relationship between sensati on seeking and age of se xual debut does exists. This null hypothesis was rejected. This findi ng corroborates those of prior studies relating sexual behavior and propensity for sens ation seeking (e.g., Donohew, Zimmerman, Cupp, Novak, Colon, & Abell, 2000; Hoyle, Fejfar, & Mill er, 2000). It is inte resting to note that sensation seeking has not before been linked specif ically to age of sexual debut; previous studies have instead tended to focus on risky sexual activ ities, such as incons istent condom use and having multiple sexual partners. The fourth, and final, set of research questi ons (RQ 4 and RQ 4.1) examined differences in ecological variables among athlete status groups.
80 Research Question 4: Do ecological variable s differ between athlet es and nonathletes? Null Hypothesis: Ecological variables do not differ between athletes and nonathletes. Relationships between athlete status and each of the six ecological variables indices, as well as the two remaining individual items were explored usi ng ANOVA. Significant differences among athlete status groups were found in mean C onnection to School and Others scores, with current athletes reporting the highest mean sc ore and nonathletes reporting the lowest mean score. This null hypothesis was re jected. Although the particular measure of Connection to School and Others ha s not previously been linked to athlete status, this finding is consistent with related studies. Egloff and Gr uhn (1996) found that athletes demonstrated higher levels of extraversion than nonathletes. Additio nally, outstanding athletes, defined as athletes spending eleven or more hours per week involved in sport, were more extraverted than lowerlevel athletes. A recent study al so found that athletes social motivation helped to explain adolescents interest in joining a sport team (A llen, 2003), suggesting that at hletes might be more outgoing and socially motivated than their nonathlete peers. Research Question 4.1: Does propensity for sens ation seeking differ between athletes and nonathletes? Null Hypothesis: Propensity for sensation seek ing does not differ between athletes and nonathletes. A t-test conducted to examine the proposed relationship between propensity for sensation seeking and age of sexual debut revealed si gnificant differences between athletes and nonathletes. The mean total BSSS score was lowe r for athletes than for nonathletes, suggesting that athletes have a higher prope nsity for sensation seeking. This null hypothesis was rejected. The conclusion that athletes, as a group, are higher sensation seeker s is consistent with that of
81 numerous previous studies (e.g., Hartman & Rawson, 1992; OSullivan, Zuckerman, & Kraft, 1998; Schroth, 1995). Sexual Activity The purpose of this study was to examine the ef fects of athletic par ticipation on adolescent sexual debut. Thus, most of the research que stions (Research Questi ons 1-3, along with the corresponding subquestions) propose relationships between independent va riables and age of sexual debut. In reviewing th e literature regarding adolesce nt sexual behavior, however, it became apparent that though age of sexual debut is an extremely important and interesting area of study, it is not the only consideration. To really attempt to unders tand the sexual decision making of adolescents and emerging adults, and the factors that most heavily influence them in making these choices, overall sexual e xperience must also be examined. Moreover, although age of debut and sexual activity are both part of an individuals overall sexual experience, they are not necessarily direc tly related; an indivi dual who delayed sexual debut until college and then became very sexually active would receive a high score. Likewise, someone who had sex for the first time in early adolescence and regrette d that decision might choose to abstain from sexual activit y in later years and, thus, receiv e a low sexual activity score. Further, the average age of sexua l debut for this sample was 17 years old; this suggests that adolescents are facing the decision of whether or not to have sex in adolescence. Thus, the items relating to factors relatin g to the high school context were lik ely a stronger influence on age of sexual debut than they might be on total sexua l activity. To further explore these possibly different effects, each of the proposed relations hips related to age of sexual debut was also explored in relation to sexual activity, as meas ured by score on the se xual activity scale.
82 Athletic Participation The primary focus of this study was the hypothesized relationship between athletic participation and sexual behavior. To explor e this relationship, ANOVA was used to see if reported sexual activity differed by athlete status. No significa nt differences were found Next, ANOVA and t-tests were c onducted to look for differences in sexual activity by intensity of sport particip ation, level of athletic identity, type of sport, level of contact, and team vs. individual sport variables. No significant differences were found. The findings from these bivariate analyses might suggest that the proposed relationships between athletic participation factors and sexual behavior did not exist. The multivariate analyses, however, provide evidence that athletic participation may have a role in an individuals overall le vel of sexual activity. Several of the sport participation variables were found to be significant in the reduced multiple regression model conducted only for athletes and including all demographic, ecological, sensation seeking, and athletic participation variables. The previous model, not including athletic participation variables, explained approximately 30% of the variance in sexual activity. Adding intensity, level of contact, team vs. individua l sport, and athletic identity to the analysis increased the models explanatory power to almost 37%. In this final reduced model, contact sport, team sport, and AIMS Social Identity we re found to be significantly and negatively related to sexual activity, while the statement I was pa rt of the most popular group at school was positively related to sexual activity. These varied findings suggest that the link between being an athlete and sexual behavior may not be as direct or as simple as suggested in prior studies. Instead, it may be some particular aspect(s) of the sport participation experience that mediate this relationship. The importance of mediating factors has been s uggested in previous research (Dodge & Jaccard, 2002; Lehman & Koerner, 2004).
83 Ecological Variables To test the relationship between ecologica l variables and sexu al activity, ANOVA was conducted to examine differences in sexual activity score by each of the six ecological variables indices and the two individual ecological variab les statements. Significant differences were found for Risk Avoidance, Values, and Parental Monitoring, with higher scores on each of these indices corresponding with lower sexual activity sc ores. The importance of these variables was further established by the multiple regression anal yses. The second sexual activity model, which included only the ecological vari ables, accounted for almost 24% of the variance in sexual activity score, suggesting a clea r link between factors in an individuals social ecology and his/her sexual activity. Furthermore, in the reduced model, four ecological variables (Risk Avoidance, Parental Monitoring, Values, and Conn ection to School and Othe rs) were found to be significant. As was the case for age of sexual debut, hi gher scores on the Risk Avoidance and Values indices were found to be protective factors. It is interesting that these ecological variables were also found to be significant for overall sexual activity, as these tw o measures of sexual experience are very different. This is especially important to note for th e Risk Avoidance index. The sample used for the sexual activity analyses di ffers from that used in the age of sexual debut analyses in that all respondents are included, even those who have not yet had sexual intercourse. The significant relationships betw een Risk Avoidance and each of these dependent variables is consistent with the findings of studies linking su bstance use to risky sexu al behaviors, including those that do not involve intercourse (e.g., Brook, Adams, Balka, Whiteman, Zhang, & Sugerman, 2004; Schuster, Bell, & Kanouse, 1996). Similar to the findings of this study, higher levels of parental monitoring have previously be en found to be associated with decreased sexual risk behavior in adolescence (Rodgers, 1999).
84 The positive relationship between Connection to School and Others and sexual activity is surprising, given the numerous studies showing that the many factors associated with school and community connection are related to a decrease in sexual risk. Douglas (2002) found that attachment to school and involvement in school programs, including programs that both did and did not involve a sexual educati on curriculum, decreased adolescen t sexual behavior. In another recent study (Tayler-Seehafer, 2000), family and school connection, as well as the presence of caring adults, were found to be protective factors against risky sexual behavior for adolescent girls. Contrary to the findings of this st udy, Small and Luster (1994) found that neighborhood monitoring was associated with decreased chan ces of being sexually experienced. Further, Mancini and Huebner (2004) found th at participation in structured time-use activities (such as sport or other extracurricular activities) was asso ciated with less risk-tak ing behavior. Perhaps the number and diversity of the items included in this index school involvement, neighborhood monitoring, presence of caring a dult, having friends at school, and enjoying going to high school contributed to the unexpected results. A dditionally, Miller, L ynam, Zimmerman, Logan, Leukefeld, and Clayton (2004) found that the extraversion personality characteristic was associated with increased sexual risk taking. Adolescents who are involved in their high school and connected to those around them might be more extraverted than their peers. Sensation Seeking An ANOVA used to examine the effect of propensity for sensation seeking on sexual activity found significant differen ces in sexual activity score by BSSS score. Those respondents who scored lower on the BSSS reported higher se xual activity scores, suggesting that higher propensity for sensation seeking wa s related to increased sexual activity. This relationship was also found in the multiple regression models. Th e third model, including only the four sensation seeking domains, accounted for over 16% of the va riance in sexual activity score, suggesting that
85 propensity for sensation seeking is an important factor in sexual decision making. In this model, the sensation seeking domains Boredom Suscep tibility and Disinhibi tion were significantly related to sexual activity. As expected, as sc ore on the Disinhibition su bscale increased, so did sexual activity score. Previous studies ha ve also found a link between the concept of disinhibition and sexual behavior (Arnett, 1990; Bancroft, Jan ssen, Carnes, Goodrich, Strong, & Long, 2004; Bogaert & Fisher, 1995). In the redu ced sexual activity mode l, Disinhibition was again found to be significantly and po sitively related to sexual activity. Interestingly, Boredom Susceptibility was nega tively related to sexual activity, with higher scores corresponding to lower sexual activity scores. This domain of sensation seeking reflects dissatisfaction with repetitive activities or routin es, and has not been tied to sexual behavior. Perhaps this finding of in the cu rrent study is reflective of the m easure used. The sexual activity score was designed to measure overall sexual activit y, but does not take into account factors such as number of partners or context (casual or ro mantic). Perhaps res pondents scoring exhibiting lower Boredom Susceptibility scores are more likely to be fulfilled and comfortable in a longterm relationship; this relationship thus provi des opportunity for frequent sexual activity. Many responses to the open-ende d question speak to the influence of being in a long-term relationship: I had sex because I felt I was ready. I was co mfortable with my partn er and didnt think it would harm our relationship. In fact, it enhanced it, we grew even closer. We had been together for awhile, and we still are. I think physical intimacy plays a large par t in ones satisfaction with ones partner, and its f un recreation if one is safe and res ponsible, so I decided to try it. Ecological Model of Sexual Debut The theoretical framework uti lized in this study was the ecological model developed by Small and Luster (1994). This model proposes that various factors at the individual, familial, and extrafamilial levels work together to influe nce an adolescents decision to become sexually
86 active. This study examined variab les at each of these levels that were identified in the literature as having an impact on adolescent sexual behavior. The efficacy of this model was supported by the findings of this study, as significant relationships between both age of sexual debut and overall sexual activity and fact ors at all three ecological leve ls were revealed. Specific relationships are discussed below. In addition to the support provided by the bi variate and multivariate analyses described thus far, this model is furthe r corroborated by partic ipant responses to th e open-ended question. Participants were asked to write their answer to an item asking what influenced their decision to have sex or to delay sex. The responses were numerous and varied, a nd existed at all three ecological levels. Selected re sponses are included in the foll owing discussion to enhance the interpretation of the results of this study within the framew ork of the ecological model. Additionally, representa tive responses, grouped by theme, can be found in Appendix E. Individual Level Numerous factors at the indivi dual level were found to infl uence age of sexual debut and overall sexual activity. For example, several important differences were found between male and female respondents. Although gender did no t appear to affect age of sexual debut, significant differences were found between male s and females in sexual activity scores. Specifically, males reported higher scores, sugge sting that they were overall more sexually active. This finding corroborates prior studies showing males report more sexual activity than their female peers (Halpern, Hallfors, Bauer, Iritani, Waller, & Hyunson, 2004; Miller, Farrell, Barnes, Melnick,& Sabo, 2005). The results of this study also suggest that the social ecologies, as measured by the ecological variables included in analysis, of males and females are experienced differently. These differences, described below, may help to explain the differing le vels of sexual activity
87 between the genders; this possibility is consistent with the results of tw o recent studies (Little & Rankin, 2001; Meschke, Zweig, Barber, & Eccles, 2000) concluding that males and females are affected by different risk and protective factors. Females reported higher mean scores on the Risk Avoidance, Parental Monitoring, and Values indices, revealing seve ral important findings. Both Risk Avoidance and Values were found to aff ect age of sexual debut, and all three of these ecological variables were significa ntly related to overall sexual activity. Generally, decreased sexual activity and older ages of sexual debut were associated with greater identification with each of these variables. The lo wer sexual activity scores reported by females may be influenced by their greater agreement with th e items within these indices. Further, females reported higher levels of agreement with the statement My schoolwork was very important to me. This statement represents a strong attachment to school and a commitment to achievement, both of which are likely to decrease sexual risk taki ng. Differing responses to this item may also contribute to the lower sexual ac tivity scores of females. Males, on the other hand, tended to receive hi gher scores on the Popularity index. In the analyses conducted for the entire sample, popularit y was not significantly related to either age of sexual debut or to overall sexual activity. In the athlete populati on, however, considering oneself to be part of the most popular group was related to increased se xual activity, suggesting that this factor is important to consider. Males who are ac tively seeking to enhance their popularity may increase their sexual activity as one means to do so. Further, males who actually are viewed as being more popular probably have greater opportunity to engage in sexual behaviors and, thus, may report higher sexual activ ity scores than their less popular peers. Additionally, Meschke, Zwieg, Barber, and Eccl es (2000) found that sexually experienced adolescents valued popularity more than their se xually inexperienced peers. Lastly, propensity
88 for sensation seeking was higher for males than fo r females. This variable, too, was related to sexual activity; higher propensity for sensat ion seeking corresponded with increased sexual activity. The higher sexual activity scores of ma les, therefore, may be partially due to their greater propensity fo r sensation seeking. The many differences found between males and fe males in this study are consistent with the findings of previous research. For exampl e, although parental monitoring is protective for both males and females, studies sh ow that parents tend to monitor girls more than boys and that this distinction may result in differential outcome s, including adolescent adjustment and problem behaviors (Jacobsen & Crockett, 2000; Richar ds, Miller, ODonnell, Wasserman, & Colder, 2004). There is also some evidence to suggest that parents have different expectations of their daughters and sons regarding sexua l activity (Moore & Rosenthal, 1991). Females, both adults and adolescents, have consisten tly been found to report higher le vels of religio sity (Maselko & Kubzansky, 2006; Smith, Denton, Faris, & Regnerus 2002). Furthermore, religiosity has been found to affect the sexual beha vior of female adolescents much more than that of male adolescents (Rostosky, Wilcox, Comer-Wright, & Ra ndall, 2004). The higher scores received by female respondents on the Risk Avoidance index al so corroborates prior research. Adolescent males are more likely than adolescent fema les to report substance use (Young, 2002). A recent study also suggests that the risk and protective factors affecti ng adolescent substance use are different for males and females (Husler & Plauch erel, 2006). Finally, ma les have consistently been classified as higher sensat ion seekers than females (Arnold, Fletcher, & Farrow, 2002; Ball, Farnill, & Wangeman, 1984; Scourfiel d, Stevens, & Merikangas, 1996). One concept that has been suggested in the liter ature to account for so me of the distinction between males and females is gender role soci alization. This is the process whereby
89 children acquire and internalize the values, attitu des, and behaviors associated with femininity, masculinity, or both (ONeil, 1995, p. 203), and is a ffected by parents, teachers, peers, media, and society at large. The intern alization of these cultural gender norms may affect the behavior of adolescents and emerging adults, as well as that of their parents. Identification with gender role stereotypes have been found to have a ne gative affect on adoles cents binge eating and problem drinking behaviors (Williams & Ricciar delli, 2003), aggression (Crothers, Field, & Kolbert, 2005), and anxiety (Palapattu, Kingery, & Ginsburg, 2006). Perhaps the higher levels of parental monitoring and stri cter values toward sex expres sed for adolescent daughters is reflective of the respective roles that males and females are expect ed to play. Females are not traditionally supposed to desire sexual experien ces and, thus, have to be protected; males, however, must be allowed some outlet for their i ndependence, and particul arly for their sexual needs. Several studies have documented the e ffects of gender role i ssues on sexual behavior, including safer sexual behavior (Buysse & Van Oost, 1997) a nd consent to unwanted sexual activity (Walker, 1997). Additiona lly, Lamb (1997) argues for the inclusion of moral education, including a discussion about gender role s, in sexual education programs. Contrary to the findings of previous studies (Miller, Fa rrell, Barnes, Melnick, & Sabo, 2005; Perkins, Luster, Villarruel, & Small, 1998; Upchurch, Levy-Storms, Sucoff, & Aneshensel, 1998), race did not seem to affect sexual behavior, as neither the bivariate nor the multivariate analyses revealed any significant di fferences in age of sexual debut or sexual activity by race group. Results from this study, how ever, do suggest that athletes of different racial backgrounds experience thei r athlete status differently. Specifically, scores on the total AIMS, Social Identity, and Nega tive Affectivity scales differed significantly by race group, with White respondents and Black respondents showing the largest distinction. For all three of the
90 scales, White athletes reported the lowest scores and Black athletes report ed the highest scores, with scores from the Hispanic and Other athletes in between. This suggests that White athletes identify more strongly with their athlete status. Furthermore, White athletes, more than athletes in other race groups, seemed to feel a stronger social identification as an athlete, identifying with the statements I considered myself an athlete; I had many goals related to sport; and Most of my friends were athletes. This strong athletic so cial identity was found in the multiple regression analysis to be related to lower sexual activity scores. Finally, White athletes report greater negative feelings when faced with sport-relate d challenges such as a poor performance or a serious injury. These findings ar e consistent with previous re search suggesting distinctions among athletes of different races (Elling & Knoppers, 2005; Miller, Melnick, & Barnes, 2005). This study found that individuals who avoided substance use delayed sexual debut longer than individuals who did not. Choosing not to smoke, drink, or use drugs might reflect an overall commitment to a healthy and responsible lif estyle and a desire to not participate in risky behaviors, which might lead an adolescent to choose not to have sex. Further, the use of alcohol impairs judgment, and can lead people to ma ke riskier choices re garding sexual activity, including sexual debut. For example, one ma le respondent describes the circumstances surrounding his sexual debut: I was drunk and determined to lose my virginity and this girl was all over me and one thing led to another. Another respondent, a female, clearly acknowledges the influence of alcohol in her sexual debut: The first time [I had sex] I was intoxicated with an older guy. Adolescents who make a point to avoid using alcohol eliminate their susceptibility to this type of ris ky sexual choice. Substance use was also found to be an importa nt influence on overall sexual behavior. Responses to the open-ended question suggest a cl ear relationship betwee n alcohol use and total
91 sexual activity: Unfortunately, most of my decisions to have sex in college have been friends of mine while under the influence of alcohol. This association might be even stronger than that for age of sexual debut, due to the sp ectrum of activities included in the sexual activity scale. While having sex for the first time under the influence of alcohol might be a major life event, perhaps kissing someone while intoxicated would not be quite as significant a choice. The finding that respondents who report high er agreement with statements reflecting religiosity have their sexual debut at a later age is not w holly unexpected. Although not all studies examining religiosity and sexual behavior have found this directional association, it is often cited as a protective f actor for adolescents. Many re spondents used the open-ended question to discuss their religious convictions and the influence of those beliefs on their sexual debut, or lack thereof. Some re presentative examples follow: Im a firm believer in Jesus Christ, therefore I choose to go by the word of God, which does not condone premarital sex. Pleasing God overrides pleasing my body ; My faith as a Christian has led me to abstain until marriage; My relationship with Christ and my strong moral convictions [influenced me]. As all respondents who wrote about the importan ce of their religious be liefs to their sexual activity identified themselves as Christians, it is importa nt to note that no difference was found in age of sexual debut or in sexual activity score between Christian respondents and those in the Other category (including respondents identif ying with another relig ious affiliation and respondents who chose No Affiliation in response to the religious affiliation item). Furthermore, no differences were found between the Christian denomination groups. This suggests that the importance of an individuals religi on to his/her life and the extent to which he/she adheres to the convictions of that religion, a nd not mere membership in a specific religious group, are the significant factors in this relationship.
92 Another important point when considering th e association between the Values index and age of sexual debut is that only those respondents who had al ready experienced their sexual debut were included in this an alysis, thus excluding any respo ndents who have abstained from sexual intercourse for religious reasons. Despite the exclusio n of these participants, higher scores on the Values index were still associated with delayed sexual debut. There are several possible reasons for this finding. First, studies designed to determine the eff ectiveness of various se xual education programs have found that faith-based abstinence-only educa tion, as well as virginity pledge, approaches generally do not prevent adolesce nts from having sex. Instead, stude nts who have participated in such programs tend to initiate sex at a slightly la ter age. One note regarding such programs is that although a later age of sexua l debut is considered a positive effect of most sexual education efforts, these specific approaches have been criticized rather than commended because adolescents emerge from these programs with no safe sex information or resources and, thus, are actually more at risk than students rece iving comprehensive sexua l education (Bruckner & Bearman, 2005). This study did not ask responden ts about their exposure to sexual education, their knowledge of safe sex, or thei r contraceptive use; therefore th e extent to which this factor might have affected the results cannot be determined. Additionally, several par ticipants indicated in the open-ende d responses that they had very strong religious convictions and intended to wait until marriage for sexual intercourse but decided to have sex before th is point for various reasons: I tried to delay sex as much as possible mostly because of my religious views. My decision to have sex was influenced because I fell in love and wanted to experience it; Hormones, feelings, pleasure. I believed premarital sex is wrong, but when you are in the moment with someone who does not agree, it is hard to
93 resist; I wanted to wait until marriage but once I got to college I changed my mind. These respondents scored very high on the Values inde x, had a later age of se xual debut, and were included in analysis, thus c ontributing to this finding. Religiosity was also associated with overall sexual activity. Although the effect of this factor on sexual activity is likely similar to that on age of sexual debut, they are distinct measures and, thus, may be influenced somewhat uniquely. First, the spectrum of behaviors included in the sexual activity scale is an important cons ideration. Though many people choose to abstain from sexual intercourse until marriage due to reli gious reasons, not all of these individuals extend this commitment to all sexual behavior s. Thus, respondents who may not yet have experienced their sexual debut woul d likely still record a sexual activity score. Furthermore, all of the sexual behaviors on the spectrum were f ound to be highly relate d; thus, respondents who choose to abstain from sexual inte rcourse due to religious convicti ons may be more likely to also limit or abstain from engaging in the other sexu al behaviors as well. The following responses, from participants who indicated very low involvement in any sexual behaviors, are representative of such feelings: In my religion it is meant fo r a married couple to engage in such acts so I am waiting for marriage; Being a Christian strongly in fluences every aspect of my life. I am not proud of what I did sexually (touching) in my past relationship but I have decided to wait until marriage to go any further than kissing from his point on. It is also possible that respondents who delayed sexual debut in adolescence for reasons related to items included in the Values index and became sexually active at a later age would have lower sexual activity scores because they would have had less time and opportunity than someone who began having sex at an earlier age.
94 Though the null hypothesis was rejected, the re lationship between sensation seeking and age of sexual debut was not a str ong one. This is possibly due to the social importance, or lack thereof, afforded to sexual debut. Though having sex for the first time can be risky and has been identified by researchers as a hi ghly important marker of adoles cent sexual behavior, it is often considered dangerous by the general population only when it happens at a very young age. This contrasts to other identified sexually risky beha viors, such as lack of contraceptive use or multiple partners; these are considered risky whet her the individual in question is 14 or 45 years old. Furthermore, although the ac t is certainly significant for t hose involved, it may not be given much thought by todays society. Unlike many othe r risk behaviors iden tified in adolescence, sex is an act that, in adulthood, is accepted and ev en encouraged; it is assumed that everyone will have sex at some point. Thus, while propensity for sensation seek ing has been related to sexual behavior, age of sexual debut is pr obably not the best test of this relationship. Sensation seeking might be more usefully examined in relati on to risky sexual behaviors or lifetime sexual experience. In fact, a positive relationship was found between high sensation seeking and increased sexual activity. Sensation seeking is defined by Zuckerman as the need for varied, novel, and complex sensations and experiences and the willingness to take physical and social risks for the sake of such experiences (1979, p. 10). It is not di fficult, given this definition, to understand the association between sensation seeking and sexu al activity. Sexual experiences encompass many of the qualities sought after by high sensation seekers (variety, novelty, complexity, physical sensation); increased sexual ac tivity more partners, greater frequency, engaging in behaviors farther along the sexual behavior spectrum will further enhance these qualities. Furthermore, sexual intercourse can be consider ed a risky activity, both physically and socially. High number
95 of sexual partners, having sex under the influe nce of alcohol, and not using condoms or birth control methods all put an indivi dual at increased risk for an STI or unintended pregnancy. Engaging in other sexual behavi ors, such as oral and anal sex, does not carry the risk of pregnancy, but can still cause indi viduals to contract an STI; th ese behaviors may actually be more risky, since adolescents/emerging adults ofte n do not consider them to be sex and, thus, do not protect themselves. Finally, choosing to e ngage in sexual behaviors could affect social standing or the quality of various relationships. Shared inform ation about sexual exchanges has the potential to cause friction between children and parents, harm friendships, end romantic relationships, or damage ones reputation. The positive relationship between score on th e Disinhibition subscale and sexual activity score was significant for the sample as a whole a nd for the athletes when selected separately, even with the inclusion of the sport participation variables. This content domain is reflective of the unrestrained behavior, or lack of concern for what others think of one s behavior, component of sensation seeking. An approach to life based on this attitude can clearly be extended to sexual behavior. If an individual doe s not care about societal norms for expected sexual behavior or does not practice restraint in his/ her sexual behavior, it would be expected that his/her sexual activity score would be rather high. Furthermor e, one item in this subscale relates to the attendance of wild parties; regularly being a pa rt of such social events may go hand-in-hand with increased or risky sexual behavior. Alt hough Disinhibition was not associated with sexual activity for nonathletes, a positiv e relationship was found for the domain of Experience Seeking. Someone with a personality that tends toward new experiences in general would probably seek out varied sexual experiences as well; such a person would likely report a high sexual activity score. It is interesting that this link was only found for the nonathletes. Perhaps athletes
96 participation in sport satisfies this particular domain, while nonathletes are motivated toward other experiences, such as sexual activity. This study corroborated an already establis hed link between propensity for sensation seeking and athlete status. It can easily be s een how an individual whose personality reflects such a drive would be attracted to sport participat ion. In any sport, even with strict rules and regulations, each game or practic e is unique, with new competitors, varying conditions of the field or the equipment, and differe nt judges or officials; variation in these factors contribute to the variety and novelty of sport pa rticipation. Furthermore, athlet es are always pushing to be better; they work to improve th eir technique, acquire new skills, and achieve the next level of competition, taking on gradually more complex goals and tasks. Lastly, athletes put themselves at risk for serious injury each time they partic ipate in their sport. Acknowledging this danger and still choosing to engage in an activity is a definite characteristic of a sensation seeking personality. Likewise, some athl etes may not even recognize thei r sport participation as a risky behavior; this underestimation of risk is also part of sensation seeking. Familial Level Several factors at the familial level were f ound in this study to be important to sexual decision making. The final item in the Valu es index, My parents would have strongly disapproved of me having sex while in high school, represents parental values toward sexual activity. Regardless of religious affiliation or degree of religi ous conviction, parents can have a strong influence on adolescents sexu al behavior. Adolescents who are close to and respect their parents and who value their pare nts opinions would likely avoid behaviors that would upset or disappoint their parents. Furthermore, children and adolescents determine which values are most important in their lives based on experience and the influence of those closest to them. The family, and particularly parents, may have the largest impact on this part of development. Thus,
97 parents who place a high value on abstaining from or delaying sexual activity are likely to have adolescents who internalize these values and, thus, delay their sexual debut. Moreover, adolescents whose parents expre ssed strong disapproval of sexual activity may have been afraid to go against these wishes. Many respondents answers to the open-ended questions reflect the importance of their parents attitudes toward sex in their decision to delay se xual debut. Although parental expectations affected the following three responde nts differently (fear, closeness, etc.), the outcome, delayed sexual debut, was the same: My parents are very religious. They would have been very angry to find out about my sexual experi ences in high school and college. I was afraid they would find out, which was the main reason to delay; I think talking to my mom about sex (we watched the miracle of life video) influenced me to wait to have sex; I grew up in a very strict household so my parents would have been very disappointed in me. Lastly, although parents attitudes toward sex probably ma y a stronger effect on younger children and adolescents, lifetime values are learned from the family. Thus, individuals who feel that their parents would have strongly disapproved of them having sex in high school might be more likely to delay or limit all sexual activity even into early adulthood. Parental monitoring probably influences sexual activity in several ways If parents know where their adolescents are going an d who they are with, they are in a better position to intervene if necessary. This could be as innocuous as ta lking about the appeal of certain friends or as significant as driving to pick up a te en at a party so he/she is not in the position of having to be in a car with someone who has been drinking. Further, if adoles cents know that they will be expected to discuss the circumst ances surrounding where they are going with their parents before they leave and when they come home, they ma y be less likely to make risky choices. These
98 factors may decrease adolescents opportunity fo r sexual activity and, additionally, help them to fully think through a situation before maki ng decisions, including those regarding sexual behavior. Finally, parents who make a purposef ul effort to monitor their adolescents whereabouts and companions may be more likely to be involved in their lives in other ways as well. Having a parent who is an always pres ent source of support might help adolescents and young adults engage in hea lthy sexual decision making. Extrafamilial Level Lastly, multiple factors at the extrafamilial level were found to impact sexual debut and overall sexual activity. Throughout the lifespan, but particularly in adolescence, the peer group is important in setting norms for both beliefs and actions. Adoles cents surrounded by close friends who believe strongly that substance use and sexual activity is not appropriate in high school and, therefore, avoid those behaviors, may internalize these beliefs and act in the same way. Likewise, an individual w hose close friends do choose to ha ve sex and drink or smoke in high school are likely to mirror th ese behaviors. The following tw o responses highlight the often negative effect of having close friends who are sexually active: I spent two years (grade 7-8) in a homeschool system. Upon reaching 9th grade, my friends talked about sex and I felt left out so when I had my first girlfriend, I rushed into it; All of my friends had done it very young so I thought I should too. I did it twi ce and realized I wasnt ready. Another possible explanation for the influence of close friends in the Risk A voidance items is the tendency to seek out peers with similar beliefs. Thus, adolescents who choos e not to have sex or drink while in high school might be more likely to become close friends with other adolescents who share these beliefs, goals, and expectations. Though the tendency to seek out peers with sim ilar values likely accounts for some of the effect of the Values scales on both age of sexua l debut and overall sexua l activity, the impact of
99 the choices and behaviors of close friends might also be reversed fo r total sexual activity; this is again attributed to the broad range of included be haviors, as well as the longitudinal nature of this measure. People whose close friends are hi ghly sexually active, and who watch their friends make risky or emotionally upsetting sexual choices might use this experience to inform their own decisions and, therefore, limit th eir own sexual activity. Severa l responses to the open-ended question support this idea: I also saw how slutty my best friends became and how guys only talk to them because they had sex and I did not want to be like that; I have a lot of friends who have had sex with their boyfriends, and after th ey break up they are severely emotionally traumatized. This is also a reason why I am choosing to wait until I can ensure that my relationship will last. The stand-alone ecological values statement I felt like I received a good education in high school reflects the degree of attachment to sc hool. The feeling of receiving a good education is dependent on a variety of factor s. Clearly, the actual quality of the school environment, the teachers, and the available resources impacts th e overall quality of the education received by students; these are beyond the control of th e adolescent. Attitude toward school and commitment to achievement, however, are person al characteristics. Respondents who agreed more strongly with this statement probably en joyed going to school more and worked hard to succeed both in and out of the classroom. This type of achievement motivation and attachment to school may serve as a protec tive factor against early sexual debut; students who are strongly motivated to succeed now and in the future will li kely not want to risk this success by engaging in sexual intercourse at a young age. Interestingly, Connection to School and Others was positively related to sexual activity; a higher score on this index was related to increased sexual activity. This relationship may, at first
100 glance, seem counterintuitive. Several of the idea s encapsulated within this index were shown in the literature review to be pr otective factors regarding sexual behavior. Connection to School and Others, though, reflects involvement in school activities, feelings of enjoying school and having friends there, and being cared for a nd protected by adults at school and in the neighborhood. Respondents who report identifica tion with these concepts were most likely active participants in class, around the school, and throughou t the community, thus increasing their exposure to other people, and particularly ot her adolescents with similar interests and levels of motivation. It seems likely that participants who score higher on this index also report higher sexual activity scores because they have more desi re and opportunity to meet new friends, dates, and potential sexual partners. This relationship between athlete status and f eelings of connectedness is to be expected. Athletes, just because of their sport participati on, are part of a cohesive group of people with similar interests. These positive feelings ar e probably even more pronounced for athletes competing for school teams, compar ed to athletes competing for teams in the larger community, because they most likely receive greater social recognition and, thus feel connected to an even larger pool of people. Furthermore, though some athletes are committed to their primary sport to such an extent that they do not have the time, en ergy, or desire to particip ate in other activities, other athletes are members of multiple sport teams and/or also participate in other extracurricular activities, such as band, student government, or sp ecial interest clubs. In this case, athletic participation might be reflective of a personal ity type that seeks out the camaraderie and achievement opportunities provided by both sports a nd other activities. Additionally, the athletecoach relationship might contribute positively to athl etes feelings of connectedness. One of the statements included in this inde x is I had at least one adult ot her than my parents who really
101 cared about me when I was in hi gh school. In the case of student athletes, coaches or trainers might fill this role, providing adolescents on th eir teams with support, guidance, and respect. Current athlete status was found to be significantly related to sexual activity in the final reduced multiple regression model. This was a negative relationship, revealing that current athletes reported lower sexual activity scores than nonathletes. Those respondents who have remained active in sport may be strongly commit ted to their athletic participation; this commitment may, in turn, affect their attitude toward sexual behavior. Sexual intercourse, as well as other sexual behaviors on the spectrum, can put an individual at risk for pregnancy or contracting a sexually transmitted infection (STI). Either of these consequences could have a very negative effect on an athletes career. The lower sexual activity score among current athletes, then, might reflect the importance of sport in their liv es and a resulting commitment to avoid factors that may distract from or harm thei r athlete status. Alternatively, current athletes may be too busy with their sport and not have as much time and, thus, opportunity for sexual relationships. Additionally, several sport partic ipation variables were found to be significantly related to sexual activity in the final reduced multiple regr ession model for athletes. Respondents playing contact and team sports reported lower sexual activity scores than their noncontact and individual sport counterparts. It is not clear why these associations should exist. The opposite relationship between level of contact and sexual activity might actually be expected. Perhaps, however, contact sport players are sufficiently stimu lated on the field and, therefore, do not seek out sexual activity to fill this role. One po ssible explanation for th e relationship between increased sexual activity and individual sport is the greater likelihood that participants of these sports would be on coeducational teams. Swim ming/diving, for example, is a popular individual
102 sport in which there are genera lly both males and females on the same team. Being teammates with members of the opposite sex who share a nd understand the same sport-related goals and interests might be the basis for some sexual relati onships. In fact, several athletes in this study addressed the effect of having te ammates of the opposite sex. A male swimmer had this to say: I practice a sport that is surrounded by halfnaked girls, athletic and mostly good looking, and sex isnt a big deal. A female respondent on the crew team also discussed close relationships with team members. My crew team was co-ed and we were all very close. I dated a guy on the team and all my good friends were on the team. These people are still my good friends and Im still dating the guy. Athletes reporting higher Social Identity scor es received lower scores on the sexual activity scale, while athletes who considered themselves to be part of the most popular group in school reported higher sexual activity scores. The oppos ing influences of the Social Identity and popular crowd membership variables suggests that although popularity may be a risk factor for athletes, strong social identification with their at hlete status may actually be protective. Perhaps athletes who are strongly attached to their athlete status and who va lue this status as their social representation to others are careful to preserve it, thus avoi ding activities that might risk their athlete standing. For many individuals, sexual ac tivity might fall into this category of risky behaviors to avoid. Contracting an STI or b ecoming/getting someone pre gnant could potentially end or, at the very least, disrupt an athletes sport career. Furt hermore, a high school is a small community and even very personal information often becomes common knowledge. Information about an athletes sexual endeavor s may be harmful to his/her social standing as a serious athlete. Those individuals, on the other hand, for whom popularity is most important may actually benefit from this open social environment and th e social standing afforded athletes; this might
103 allow for increased opportunity and, thus, higher levels of sexual activity. These findings may be related to the distinction be tween sport participation (athlete status, social identity) and identification with the jo ck label (self-reported membership in most popular group). In previous studies, being a jock was associated with a va riety of risky behavior s, including risky sexual practices, while being an athlete actually was a pr otective factor (La Greca, Prinstein, & Fetter, 2001; Miller, Farrell, Barnes, Melnick, & Sabo, 2005). Contribution to the Literature This study added to the litera ture on the ecological model of sexual debut in several important ways. First, the original model was deve loped as a risk-factor a pproach. While this is an often successful and highly informative a pproach, it may ignore possible protective factors that could affect sexual debut. This study consid ered both risk and prot ective factors, thus broadening the models scope. Secondly, this m odel was originally developed to differentiate between adolescents who had experienced their sexual debut and those who had not. This study, however, extended its use to expl ain age of sexual debut and overa ll adolescent sexual activity. This was an appropriate extens ion, supported by the significant re sults, as many of the factors affecting sexual debut also a ffect other sexual behaviors. Additionally, this study introduced a new concep t, sport participation, into the model. Sport participation was an appropriate addition to the model, as it fit both into and around the various levels of ecology. Sport is a social activity occurring outside of the family; thus, it fits into the extrafamilial level of this model. It also, however, encompasses and affects many factors at the individual and familial levels. Th e results of this study, for example, suggest a strong relationship between athlet e status and propensity for se nsation seeking, which is an individual factor influe ncing sexual behavior. In addition, parental mon itoring (a family-level factor) was significantly related to sexual activity for nonathletes, but not for athletes. Lastly,
104 athletes reported higher scor es on the Connection to School and Others index, which was positively related to sexual activity. Finally, the data analysis methodology utilized in this study allowed for the corroboration of the underlying assumption of this theory, namely that it is not one factor but the interactions among numerous factors, that determines when an individual will have sex for the first time. In this study, bivariate analyses te sting the relationship between ag e of sexual debut and a single predictor variable produced very few significant results. However, the reduced multiple regression model, which included a number of very different variab les and is, therefore, a more accurate approximation of the conditions under which an adolescent might have to make this difficult decision, accounted for over 15% of the va riance in age of sexual debut. These findings provide support for the conclusion that a combin ation of important factors influences sexual debut much more than do any of those factors individually. This study also represents contributions to th e literature regarding sport participation and adolescent sexual behavior. First, although sport participation has been examined for its effect on various indicators of sexual be havior, the two measures used in this study (age of sexual debut and total sexual activity) have not specifically been explored before. The results of this study, then, broaden the conclusions reached through previous research. Secondly, the relationship between sexual behavi or and sport participation has not previously been considered within an ecological context, as presented in th is study. The real-world approach of the multiple regression models provid es evidence suggesting that sexual behavior is not determined by one factor, but by the combination and interaction of numerous factors. Sport participation is reflected in this approach in two ways. Athlete status was found in this study to be related to many of those factors significant to sexual activ ity (demographic characteristics, ecological
105 variables, propensity for sensati on seeking); additionally, sport par ticipation may itself be one of those important variables. Limitations and Delimitations One possible limitation to this study is the inclusion of retrospective questions. Participants were asked to answ er questions about their sexual activity, sports involvement, and other life experiences during adolescence, which may lead to memory recall error. The crosssectional design of this study is another lim itation. Although correlations and relationships between the variables can be id entified, causation cannot be esta blished. Generalizability is a delimitation. The population in this study was unde rgraduate students from a large southeastern university; therefore, results should only be generalized to similar populations. The proportion of female respondents in this sample, much larger than that of females in the University as a whole, may also be a delimitation. Implications for the Future Research A thorough review of the litera ture and examination of the results of this study have identified several areas for future research. Firs t, the ecological model of sexual debut identifies four levels of influence: individual, familial, extrafamilial, and macrosystem. This study only explored factors at the first three levels. T hus, future research should examine macrosystem factors that might affect adol escent sexual behavior. Examples of possible topics for study include the media, the sexual culture of the larg er society, and federal or state sexual education policies. Secondly, this was an exploratory study attempting to identify possibl e relationships between numerous sport particip ation variables and sexual activit y. Narrowly focused research with specifically targeted athlete samples ma y provide more comprehensive results. For example, one relationship that this study was unabl e to clarify was the association between level
106 of contact and sexual behavior. This was due to the small number of collision sport participants, as well as the smaller proportion of males, who were more likel y to play both collision and contact sports, in the study samp le. Targeting larger numbers of these athletes and comparing them to the noncontact athletes may help elucidate this relationship. Thirdly, although the results of this study were able to shed some light on factors influencing age of sexual debut, the findings ma y be somewhat impacted by the retrospective nature of the study. Thus, future research shou ld attempt to study sexual debut with samples of current adolescents, as these young individuals are likely to be more immediately dealing with this important decision; however, it may be difficult to get permi ssion to talk to minors about sex. Lastly, in response to the open-ended que stion, many participants suggested that, in addition to sport participation, ot her extracurricular activities might also affect sexual behavior. For example, one respondent offered the following suggestion; Sports activities are certainly a valid connection to intimate relationships the arts offer the same connections too maybe study this participation as well? Thus, future research should include both athletics and other activities. This would serve two purposes. Firs t, it would allow for the teasing apart of the specific effects of sport participation and the e ffects of involvement in general; and second, it would allow researchers to compare students who choose to participate in sports and those who choose other activities. Practice The results of this study provi ded evidence that athletes and nonathletes are two distinct groups of students. Understanding this distin ction may help school officials and youth workers more effectively reach both populations, specifi cally in the realm of sexual behavior. The suggestions of Schaalma, Abraham, Gillmore, and Kok (2004) support this idea. They state that in determining the most effective approach to sex education, it must be remembered that
107 different target populations ma y necessitate different interv entions or combinations of interventions. In this study, a direct link be tween athlete status and sexual behavior was not found; instead, possible differences in sexual ac tivity between athletes and nonathletes were revealed through various mediating variables. Athletes, as a group, demonstrate a higher propensity for sensation seeking, which was rela ted to increased sexual activity. Recognizing this fact allows practitio ners working directly with youth to ta rget students involved in sport, and to ensure that this drive for excitement is focu sed in healthy, structured activities rather than risky sexual behaviors or other ri sk behaviors, such as drug or al cohol use. Coaches, teachers, or other practitioners working w ith the student athlete populatio n should emphasize the importance of sport participation and the pride they should ha ve in their achievements. This emphasis could enhance some of the positive facets of athletic identity, which were associated with lower sexual activity scores. Additionally, it is not necessa ry for programs aimed at decreasing adolescent sexual behavior in the student populati on as a whole to solely focus on or discuss sex. In fact, after reviewing trends in adolescent pregnancy, as well as societal changes that may have affected these trends, Brindis (2006) advocated for the wider implementation of youth development programs. Such programs, with an ecological pe rspective as their base and asset-building as their goal, should promote psycho logical strength (resiliency) and prosocial behaviors. Key factors are programs and committed individuals to provide the nurturing and attention necessary for young people to feel that making socially positi ve decisions can truly make a difference (p. 287). The author goes on to suggest some possi ble components of this type of program: job readiness training, academic tutoring, life ski lls training, peer teaching or counseling, and recreation, including particip ation in sports. The results of th is study support th e basis of such
108 programs, particularly the acknow ledgment that adolescents are influenced by factors at all levels of the social ecology. Further, the im portance of overall healt hy decision making (i.e., avoiding substance use), positive peer influence, and school attachment, all factors Brindis says should be included in these program s, was highlighted in this study. Public Policy In response to what they see as a critical lack of social science and public health research in public policy development, Bleakley and Ellis (2003) urge researchers, especially those who study adolescents, to be become more involve d. The physical, emotional, and social vulnerability of adolescence makes policy an especially crucial tool in shaping this phase of their life course (p. 1802). Although th is study did not examine any specific policies, it may be helpful to individuals in the public policy sphere. The results of this study, examining adolescent sexual behavior, can be used to inform decisions on various public policy issues. For example, rather than simply quantify sexual behaviors a nd their rates, this study attempted to identify those factors that might affect sexual decision making and which, therefore, should be addressed in programmatic efforts to decrease adolescent sexual behavior. These underlying factors could also be utilized in the developm ent of policy. Brindis (2006) argue s that, the challenge is to consider a number of policy op tions that respond not only to observable behaviors, such as adolescent childbearing, but also to their underlying antecedents (p. 281). As an example, the finding that lower degree of agreement with st atements reflecting avoidance of substance use was associated with increased sexual activity and younger age of sexual debut could be used to justify the inclusion of an alc ohol education segment in an overa ll youth program, with the final goal being the reduction of a dolescent sexual behavior. Moreover, this study found support for an ecologi cal model of adolescent sexual behavior. Neither age of sexual debut nor overall level of sexual activity was heavily influenced by any
109 one variable; instead, both meas ures of sexual behavior were impacted by a combination of variables and the interactions be tween them. In their discussion of sexual education, Schaalma et al. (2004) suggest that policies providing for se xual education avoid the traditional risk-factor approach, as well as the tendency to focus only on the presentation of facts on sexual health. Instead, these authors advocate for policies that take into account th e environmental influences of adolescent sexual decision making and provide for sexual educati on that emphasizes healthy and supportive environments. Summary Sexual behavior in adolescence and particularly sexual debut is an often controversial topic, guaranteed to cause debate among parents, school officials, politicians, policymakers, and adolescents themselves. The most effective way to handle the controversy surrounding this issue is to try to understand it more fully, to try to di scover more about it so th at better decisions can be made. This study examined the effects of at hletic participation on ad olescent sexual debut. Results of this study support an ecological approach to adolescent sexual behavior, one in which numerous factors combine and interact to influe nce an adolescents decision to have sex for the first time. Conclusions reached through this study have far-reaching implications for youth researchers, practitioners work ing with adolescents, and public policy professionals. Although this study offers a glimpse into the sexual world of adolescents, it is only th at, a glimpse. It is the sincere hope of the researcher that these findings will inspire other researchers, as well as policymakers and those directly i nvolved with youth, to delve deep er into this highly important field of study together and tr y to really make a difference in the lives of adolescents.
110 APPENDIX A CONSENT FORM ATHLETIC PARTICIPATION AND SEXUAL ACTIVITY QUESTIONNAIRE Consent Form: Thesis Research on Athletic Participation and Sexual Debut I am a graduate student in the Department of Family, Youth and Co mmunity Sciences at the University of Florida. I am conducting rese arch on the effects of certain life experiences, particularly being an athlete, on individuals sexual activity. I am specifically interested in timing of sexual debut, which is when an individua l voluntarily has sex for the first time. The purpose of this study is to examine the relations hip between sports part icipation and timing of sexual debut. With your permission, I would lik e to ask you to volunteer for this research. For this study you will be asked to comple te, on your own, an instrument containing 70 questions. This instrument will take approxi mately 20 minutes to complete. Since your participation is voluntary, you can skip any que stions you do not wish to answer and you may withdraw consent for your part icipation at any time without consequence. If you choose to withdraw, your survey will be de stroyed. There are no known risk s or immediate benefits to you as a participant. No compensation is offere d for participation. All your responses will be anonymous, as not even the researcher will be able to associate responses to particular participants. Group results of this study will be available upon request at the completion of the project. If you have any questions about this research study, please contact th e student researcher at (352) 392-2202 ext. 308 or email firstname.lastname@example.org or the faculty supervisor, Dr. Rose Barnett, at (352) 392-2202 ext. 248 or email email@example.com Questions or concerns about your rights as research partic ipants may be directed to the UFIRB office, University of Florida, P.O. Box 112250, Gainesville, FL 32611, (352) 392-0433. Sincerely, Sally Moore Graduate Student Department of Family, Youth and Community Sciences I have read the procedure above. By completi ng the following survey I voluntarily give my consent to participate in the Thesis Resear ch on Athletic Participation and Sexual Debut.
111 APPENDIX B INSTRUMENTATION Directions : Please answer the following questions to the best of your ability. If you cannot answer a question you may skip it, but try to answer all questions. Circle or write your answer for each qu estion directly on this sheet. Your answers will remain anonymous For the purpose of this survey, someone would be considered to have been an a thlete in high school if he or she competed or participated in an organized sport, either on a school team or in the larger community. First, we are going to ask you some questions about your participation in sports. Please answer the first question and then follow the prompt to the next question that is applicable to you. Which of the following applies to you? 1. I was an athlete in high school and I am now a collegiate athlete competing for the University of Florida. 2. I was an athlete in high school and I now par ticipate in University of Florida club or intramural sports. 3. I was an athlete in high school and I am not currently active in sports. 4. I was not an athlete in high school. (Please skip to the next section.) If you indicated that you were an athlet e in high school, continue the survey with the next question. If you indicate d that you were not an athlete in high school, skip to the next section. In what sport(s) did you participate? _______________________________ If you participated in multiple sports, which do you consider your primary sport? ______________________________________ On average, how many hours per week did you participate in your sport (including practices, competitions, etc.)? _______ hours
112 Athletes experience their sport participatio n in different ways. Reflect back to when you were in high school. The follo wing questions relate to your sport participation in high school. Please circ le the number that best corresponds to your sport experience WHEN YOU WERE IN HIGH SCHOOL. Strongly Agree Strongly Disagree I considered myself an athlete. 1 2 3 4 5 6 7 I had many goals related to sport. 1 2 3 4 5 6 7 Most of my friends were athletes. 1 2 3 4 5 6 7 Sport was the most important part of my life. 1 2 3 4 5 6 7 I spent more time thinking about sport than anything else. 1 2 3 4 5 6 7 I felt bad about myself when I do poorly in sport. 1 2 3 4 5 6 7 I would have been very depr essed if I were injured and could not compete in sport. 1 2 3 4 5 6 7 I was part of the most popular group at school. 1 2 3 4 5 6 7 People have very different lif e experiences as they are growing up. Reflect back to when you were in high school. The following questions relate to your life WHEN YOU WERE IN HIGH SCHOOL. Pl ease indicate the degree to which you agree with the following statements. Strongly Strongly Agree Disagree I attended religious services regularly. 1 2 3 4 5 Religion was very important in my life. 1 2 3 4 5 I avoided smoking cigarettes. 1 2 3 4 5 I avoided using marijuana. 1 2 3 4 5
113 (Continued) Strongly Strongly Agree Disagree I avoided drinking alcohol. 1 2 3 4 5 My schoolwork was very important to me. 1 2 3 4 5 My parents would have strongly di sapproved of me having sex while in high school. 1 2 3 4 5 My parents always knew where I was going when I left home. 1 2 3 4 5 My parents always knew who I was with. 1 2 3 4 5 My parents always knew where I was after school. 1 2 3 4 5 I felt close to at least one parent. 1 2 3 4 5 I talked to at least one parent about sex. 1 2 3 4 5 My parents always supported me. 1 2 3 4 5 I felt like I could talk to my mother about anything. 1 2 3 4 5 I felt like I could talk to my father about anything. 1 2 3 4 5 My mother was always there for me. 1 2 3 4 5 My father was always there for me. 1 2 3 4 5 My close friends avoided smoking cigarettes. 1 2 3 4 5 My close friends avoided using marijuana. 1 2 3 4 5 My close friends avoi ded drinking alcohol. 1 2 3 4 5 My close friends starting having se x when they were in high school. 1 2 3 4 5 I tried to do things to be more popular. 1 2 3 4 5 If my friends were doing something, I wanted to do it too. 1 2 3 4 5 At my high school, it was cons idered cool to have sex. 1 2 3 4 5 I felt like I received a go od education in high school. 1 2 3 4 5
114 (Continued) Strongly Strongly Agree Disagree I liked going to high school. 1 2 3 4 5 High school sport was very important in my life. 1 2 3 4 5 I was involved in extracurricular act ivities other than sports at my high school. 1 2 3 4 5 I had lots of friends at my high school. 1 2 3 4 5 I cared about high school mostly for the social atmosphere. 1 2 3 4 5 I had at least one adult other than my parents who really cared about me when I was in high school. 1 2 3 4 5 People in my neighborhood watched out for me. 1 2 3 4 5 People have different attitudes and preferen ces regarding how they like to spend their free time. Thinking about your life NOW, please indi cate the degree to which you agree with the following statements. Strongly Strongly Agree Disagree I would like to explore strange places. 1 2 3 4 5 I get restless when I spend too much time at home. 1 2 3 4 5 I like to do frightening things. 1 2 3 4 5 I like wild parties. 1 2 3 4 5 I would like to take off on a tr ip with no pre-planned routes or timetables. 1 2 3 4 5 I prefer friends who are exciti ngly unpredictable. 1 2 3 4 5 I would like to try bungee jumping. 1 2 3 4 5 I would love to have new and exciting experiences, even if they are illegal. 1 2 3 4 5
115 People have all kinds of experiences. Now we are going to ask you some questions about your sexual experiences. Remember that this questionnaire is completely anonymous Individual respondents cannot be connected to their answers. Your responses will be included only together with all other responses. Please indicate on a scale of 1 (Never ) to 5 (A Lot) how often you have engaged in the following sexual behaviors. If you have engaged in a behavior at least once, please indicate in the space prov ided how old you were the FIRST time. For example, if you were 16 years and 4 mo nths old, you would write 16 in the space for Year and 4 in the space for Month. Never A Lot Years Months Kissing 1 2 3 4 5 Years Months French Kissing 1 2 3 4 5 Years Months Touching a partners br east or having your breast touched by a partner 1 2 3 4 5 Years Months Touching a partners pe nis or having your penis touched by a partner 1 2 3 4 5 Years Months Touching a partners vagina or having your vagina touched by a partner 1 2 3 4 5 Years Months Performing oral sex 1 2 3 4 5 Years Months Receiving oral sex 1 2 3 4 5 Years Months Sexual intercourse (vaginal penetration) 1 2 3 4 5 Years Months If you have had sexual intercourse, was your first time voluntary? 1. No 2. Yes
116 If you have had sexual intercourse how many partners have you had? ________ If you have had oral sex, how many partners have you had? _______ Finally we are going to ask a few questions about you to help us interpret your answers. Remember that this questio nnaire is completely anonymous. Answering these questions will not reveal yo ur identity. Your name will never be placed on the questionnaire or associated with your answers. These questions are included to ensure that this sample is representative of the University of Florida student population. Are you? 1. Male 2. Female What influenced your decision to have sex or to delay sex? Please use the space provided in this box to answer this question.
117 What is your race/ethnic origin? 1. Hispanic 2. Asian/Pacific Islander 3. White (non-Hispanic) 4. Black (non-Hispanic) 5. Other _________________ How old are you? ______ Years and ______ Months What is your religious affiliation? 1. Jewish 2. Muslim 3. Christian What denomination? ___________________ 4. Other _________________ 5. No affiliation In the space provided below, please feel free to offer any information that you feel would help us better understand your decisi ons, sports participation, leisure time preferences, or other life experiences. Thank you for your time. Your participation in this study is greatly appreciated.
118 APPENDIX C SCALE CORRELATION TABLES Table C-1 Pearson correlations for AIMS AIMS AIMS1 AIMS2 AIMS3 AIMS4 AIMS5 AIMS6 AIMS7 AIMS 1 .736** .860**.687**.822**.807**.734** .764** AIMS1 1 .717**.477**.472**.449**.501** .514** AIMS2 1.563**.663**.647**.576** .559** AIMS3 1.507**.510**.379** .383** AIMS4 1.798**.489** .570** AIMS5 1.515** .519** AIMS6 1 .623** AIMS7 1 ** Significant at the.01 level Table C-2 Pearson correlations for AIMS social identity Social identity AIMS1 AIMS2 AIMS3 Social identity 1 .836**.899**.801** AIMS1 1.717**.477** AIMS2 1.563** AIMS3 1 ** Significant at the.01 level Table C-3 Pearson correlations for AIMS exclusivity Exclusivity AIMS4 AIMS5 Exclusivity 1.954**.939** AIMS4 1.798** AIMS5 1 ** Significant at the.01 level Table C-4 Pearson correlations for AIMS negative affectivity Negative affectivity AIMS6 AIMS7 Negative affectivity 1.875**.917** AIMS6 1.623** AIMS7 1 ** Significant at the.01 level AIMS1 I considered myself an athlete. AIMS2 I had many goals related to sport. AIMS3 Most of my friends were athletes. AIMS4 Sport was the most important part of my life. AIMS5 I spent more time thinking about sport than anything else. AIMS6 I felt bad about myself when I did poorly in sport. AIMS7 I would have been very depressed if I were injured and could not compete in sport.
119 Table C-5 Pearson correlations for risk avoidance index RA RA1 RA2 RA3 RA4 RA5 RA6 RA7 RA 1 .615** .733**.793**.758**.847**.814** .537** RA1 1 .620**.439**.483**.340**.312** .130** RA2 1.603**.397**.564**.406** .245** RA3 1.427**.564**.665** .305** RA4 1.713**.589** .267** RA5 1.707** .380** RA6 1 .400** RA7 1 ** Significant at the.01 level RA1 I avoided smoking cigarettes. RA2 I avoided using marijuana. RA3 I avoided drinking alcohol. RA4 My close friends a voided smoking cigarettes. RA5 My close friends a voided using marijuana. RA6 My close friends a voided drinking alcohol. RA7 My close friends started having sex wh en they were in high school. (This item was not reverse-coded.) Table C-6 Pearson correlations for parental connection index PC PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC 1 .703** .676**.679**.777**.695**.689** .601** PC1 1 .422**.493**.546**.324**.510** .272** PC2 1.321**.591**.335**.300** .103* PC3 1.463**.314**.594** .375** PC4 1.406**.548** .168** PC5 1.253** .600** PC6 1 .353** PC7 1 Significant at the.05 level ** Significant at the.01 level PC1 I felt close to at least one parent. PC2 I talked to at l east one parent about sex. PC3 My parents always supported me. PC4 I felt like I could talk to my mother about anything. PC5 I felt like I could talk to my father about anything. PC6 My mother was always there for me. PC7 My father was always there for me.
120 Table C-7 Pearson correlations for parental monitoring index PM PM1 PM2 PM3 PM 1 .877**.923**.879** PM1 1.739**.623** PM2 1.747** PM3 1 ** Significant at the.01 level PM1 My parents always knew wher e I was going when I left home. PM2 My parents always knew who I was with. PM3 My parents always kne w where I was after school. Table C-8 Pearson correlations for c onnection to school and others index CON CON1 CON2 CON3 CON4 CON5 CON6 CON 1 .646** .585**.540**.596**.523** .579** CON1 1 .258**.287**.359**.198** .189** CON2 1.119*.190**.086 .133** CON3 1.346**.198** .076 CON4 1.219** .201** CON5 1 .285** CON6 1 Significant at the.05 level ** Significant at the.01 level CON1 I liked going to high school. CON2 High school sport was ve ry important in my life. CON3 I was involved in extracurricular activ ities other than sports at my high school. CON4 I had lots of friends at my high school. CON5 I had at least one adult other than my parents who really car ed about me when I was in high school. CON6 People in my ne ighborhood watched out for me. Table C-9 Pearson correlations for values index VAL VAL1 VAL2 VAL3 VAL 1 .911**.905**.696** VAL1 1.819**.421** VAL2 1.426** VAL3 1 ** Significant at the.01 level VAL1 I attended religious services regularly. VAL2 Religion was very important in my life. VAL3 My parents would have strongly disapp roved of me having sex while in high school.
121 Table C-10 Pearson correla tions for popularity index POP POP1 POP2 POP3 POP4 POP 1 .722**.739**.624** .565** POP1 1.523**.224** .224** POP2 1.278** .195** POP3 1 .142** POP4 1 ** Significant at the.01 level POP1 I tried to do things to be more popular. POP2 If my friends were doing something, I wanted to do it too. POP3 At my high school, it was considered cool to have sex. POP4 I cared about high school mostly for the social atmosphere. Table C-11 Pearson correlations for BSSS BSSS BSSS 1 BSSS 2 BSSS 3 BSSS 4 BSSS 5 BSSS 6 BSSS 7 BSSS 8 BSSS 1 .594** .423** .766**.631**.619**.657**.602** .723** BSSS 1 1 .224** .432**.201**.351**.344**.233** .267** BSSS 2 1 .261**.190**.108* .219**.099* .143** BSSS 3 1.417**.330**.372**.520** .509** BSSS 4 1.228**.377**.198** .585** BSSS 5 1.517**.222** .346** BSSS 6 1.178** .425** BSSS 7 1 .369** BSSS 8 1 Significant at the .05 level ** Significant at the .01 level BSSS1 I would like to explore strange places. BSSS2 I get restless when I spend too much time at home. BSSS3 I like to do frightening things. BSSS4 I like wild parties. BSSS5 I would like to take off on a trip w ith no pre-planned routes or timetables. BSSS6 I prefer friends who are exciting unpredictable. BSSS7 I would like to try bungee jumping. BSSS8 I would love to have new and excitin g experiences, even if they are illegal.
122 Table C-12 Pearson correlations for sexual activity scale SA SA 1 SA 2 SA 3 SA 4 SA 5 SA 6 SA 7 SA 8 SA 1 .712** .759** .901**.917**.924**.852**.889** .833** SA 1 1 .856** .684**.597**.613**.519**.531** .519** SA 2 1 .721**.645**.672**.584**.585** .532** SA 3 1.873**.880**.705**.738** .668** SA 4 1.899**.767**.807** .728** SA 5 1.739**.819** .745** SA 6 1.806** .707** SA 7 1 .774** SA 8 1 ** Significant at the .01 level SA 1 Kissing SA 2 French Kissing SA 3 Touching a partners breast or having your breast touched by a partner SA 4 Touching a partners penis or having your penis touched by a partner SA 5 Touching a partners vagina or having your vagina touched by a partner SA 6 Performing oral sex SA 7 Receiving oral sex SA 8 Sexual intercours e (vaginal penetration)
123 APPENDIX D TABLES OF SIGNIFICANT RESULTS Table D-1 T-tests for ecological variables by gender Male Female t Risk avoidance 23.18 25.75 -3.452*** Parental monitoring 10.54 11.92 -4.325*** Values 10.12 11.17 -2.667** Popularity 8.18 6.87 5.336*** My schoolwork was very important to me. 4.19 4.66 -5.569*** ** Significant at .01 level *** Significant at .001 level Table D-2 T-test for BSSS by gender Male Female t BSSS score 20.92 23.67 -4.121*** *** Significant at .001 level Table D-3 T-tests for sport pa rticipation variables by gender Male Female t AIMS 22.23 24.99 -2.442* AIMS exclusivity 8.68 9.67 -2.529* AIMS negative affectivity 6.04 6.85 -2.044* Intensity 17.12 13.97 3.209*** Significant at .05 level *** Significant at .001 level Table D-4 T-tests for ecological variables by religious affiliation Christian Other t Risk avoidance 25.68 23.07 3.328*** Connection to school and others 23.93 21.57 5.571*** Values 12.00 7.64 12.342*** *** Significant at .001 level Table D-5 T-test for BSSS by religious affiliation Christian Other t BSSS score 23.24 21.83 1.980* Significant at .05 level
124 Table D-6 ANOVA for ecological va riables by race/ethnic origin White Black Hispanic Other F Connection to school and others 23.74 23.78 23.45 20.91 7.588*** *** Significant at .001 level Table D-7 ANOVA for BSSS by race/ethnic origin White Black Hispanic Other F BSSS score 22.39 25.57 22.48 21.57 5.595*** *** Significant at .001 level Table D-8 ANOVA for sport participati on variables by race/ethnic origin White Black Hispanic Other F AIMS 22.97 27.67 23.64 27.30 4.433** AIMS social identity 7.68 9.52 8.06 9.74 3.759* AIMS negative affectivity 6.14 8.07 6.48 7.56 5.495*** Significant at .05 level ** Significant at .01 level *** Significant at .001 level Table D-9 ANOVA for age of sexual debut by risk avoidance index 1st quartile RA score 2nd quartile RA score 3rd quartile RA score 4th quartile RA score F Age of debut 16.67 17.04 17.68 18.30 8.984*** *** Significant at .001 level Table D-10 ANOVA for age of se xual debut by values index 1st quartile Val score 2nd quartile Val score 3rd quartile Val score 4th quartile Val score F Age of debut 16.60 17.23 17.58 17.60 5.862*** *** Significant at .001 level Table D-11 ANOVA for age of se xual debut by I felt like I r eceived a good education in high school ecologi cal variables individual item SD D N A SA F Age of debut 16.02 17.77 16.53 17.22 17.45 3.783** ** Significant at .01 level (SD = Strongly Disagree, D = Disagree, N = Neutral, A = Agree, SA = Strongly Agree)
125 Table D-12 T-test for sexual activity by gender Male Female t Sexual activity score 27.70 24.78 2.683** ** Significant at .01 level Table D-13 ANOVA for sexual activity by BSSS 1st quartile BSSS score 2nd quartile BSSS score 3rd quartile BSSS score 4th quartile BSSS score F Sexual activity 28.76 26.58 23.63 22.76 8.064*** *** Significant at .001 level Table D-14 ANOVA for sexual activ ity by risk avoidance index 1st quartile RA score 2nd quartile RA score 3rd quartile RA score 4th quartile RA score F Sexual activity 30.58 27.76 24.31 18.21 30.680*** *** Significant at .001 level Table D-15 ANOVA for sexual activity by parental monitoring index 1st quartile PM score 2nd quartile PM score 3rd quartile PM score 4th quartile PM score F Sexual activity 29.18 26.21 22.96 22.07 11.867*** *** Significant at .001 level Table D-16 ANOVA for sexual activity by values index 1st quartile Val score 2nd quartile Val score 3rd quartile Val score 4th quartile Val score F Sexual activity 28.72 27.09 25.51 20.77 12.305*** *** Significant at .001 level Table D-17 T-test for sexual ac tivity by gender (athletes only) Male Female t Sexual activity score 28.48 24.96 2.936** ** Significant at .01 level
126 Table D-18 ANOVA for connect ion to school and other index by athlete status Current athlete H.S. athlete Nonathlete F Connection 24.73 23.83 21.71 22.016*** *** Significant at .001 level Table D-19 T-test for BSSS by athlete status Athlete Nonathlete t BSSS score 22.17 24.31 -3.188** ** Significant at .01 level
127 APPENDIX E SELECT RESPONSE TO OPEN-ENDED QUESTIONS Individual Level Culture/Ethnicity My culture had a big influence. People from my culture dont have sex until they are married. For my own personal protection. Im also wa iting for the right person to have sex with. Im also Hispanic and that has a lot to do with why I have not participated in sex. Alcohol Use I was drunk and determined to lose my virg inity and this girl was all over me and one thing led to another. I was drunk and horny. My decisions to have sex in high school were made when I was in long-term relationships, and considered myself to have love d the person. Unfortunately, most of my decisions to have sex in college have been friends of mine while under the influence of alcohol. Fear of Consequences My mother got pregnant when she was 18, after her first time, thats what always scared me enough to never do it when I was so young and nave. My boyfriend and I have been together for two years and hes the only person I have had sex with and the only one I plan to ever have sex with. Cant afford kids or the risk. The fact that I am not ready for the chance to get a life threatening STD or a child. I dont think its worth the risks. Religiosity In my religion it is meant for a married couple to engage in such acts so I am waiting for marriage. Im a firm believer in Jesus Christ, therefor e I choose to go by the word of God, which does not condone premarital sex. Pleasing God overrides pleasing my body. I dont conform to the ways of this pa ssing world. It costs TOO MUCH! I want to be a youth pastor. Sexual purity has always been a struggle for me I even ended a long-term relationship because of it. My future and what I plan to tell my future youth group is why Im waiting. Religious beliefs skewed my perceptions on sex. Curiosity I was curious and excited. Curiosity is what influenced my decision to have sex.
128 Physical Desire Sex is good. Ummwhy does anyone have sex? Its fun. I guess hormones influenced me. Kissing led to one thing and then sex. My body was urged to do it. I wanted to try it. I wanted to bust a nut. Skeet skeet son! Personal Feelings of Readiness It just hasnt been the right time. I am waiting for the right guy. Just looking for a person that I connect with on all levels, not just physically. Familial Level Family Expectations My parents are very religious. They would have been very angry to find out about my sexual experiences in high school and into college. I was afraid they would find out, which was the main reason to delay. Probably the biggest deterrent was the fact t hat I have three extreme ly protective older brothers and a very religious father. My fe ar of disappointing th em by ruining their pure perception of me made me wait. My parents and my brother because they alwa ys stressed education and being something in this world. I wanted to experience it. I delayed it fo r so long because of parental guidance and advice. I didnt care about abstaining after I moved out of my parents home. Extrafamilial Level Peer Influence Being pressured by boyfriend. Peer pressure pressures from males and my best girl friend. To feel accepted. Long-Term Relationship I was with the girl for quite some time and I was in love! Was in a fairly long-term relationship with an older boy. Felt comfortable with him and have never regretted it. In a long-term, committed, loving relationship and felt comfortable, safe, and ready. Involvement I wasnt concerned with it unt il I got to college. I was in sports, extracurricular activities, community service I high school, so I wasnt concerned with it in high school. Sports and school really occupy my time.
129 Sport Participation Delay sex due to sports!!! Influenced due to college and friends. Being in high school sports kept me out of trouble. I practice a sport that is surrounded by halfnaked girls, athletic and most of the time good looking, and sex isnt a big deal. Sports defined me as a person. Before sports I was a nobody with different goals. Sports helped me relieve stress, build communication an d leadership skills. It was also a social avenue to discuss sex and other social issues. Within three months of my first intercourse experience I stopped playing all high school sports and spent the majority of my free ti me working. Cause and effect, I dont know; correlation, yes. Sports give structure to my life and have taught me all the lessons about conflict, pressure, failure and success, and how to deal with them. I wish I still had a team and schedule to keep my life organized. For me, sports are about meeting high goals an d expectations that you set for yourself and other people have set for you. I think bei ng highly involved in athletics increases self-worth, which makes young people (especially girls) more likely to wait until they are sure they are ready for the consequences.
130 LIST OF REFERENCES Allen, J. (2003). Social motivation in youth sport. Journal of Exercise and Sport Psychology, 25 551-567. Arnett, J. (1990). Contraceptive use, sens ation seeking, and adoles cent egocentrism. Journal of Youth and Adolescence, 19 171-180. Arnold, P., Fletcher, S., & Farrow, R. (2002). Condom use and psychologic al sensation seeking by college students. Sexual & Relationship Therapy, 17 355-365. Ball, I., Farnill, D., & Wangeman, J. (1984). Se x and age differences in sensation seeking: Some national comparisons. British Journal of Psychology, 75 257-265. Bancroft, J., Janssen, E., Carnes, L., Goodric h, D., Strong, D., & Long, J. (2004). Sexual activity and risk taking in young heterosexual me n: The relevance of sexual arousability, mood, and sensation seeking. The Journal of Sex Research, 41 181-192. Baumert, P., Henderson, J., & Thompson, N. (1998) Health risk behavi ors of adolescent participants in organized sports. Journal of Adolescent Health, 22 460-465. Beyth-Marom, R., Austin, L., Fischhoff, B., Palmgren, C., & Jacbos-Quadrel, M. (1993). Perceived consequences of risky be haviors: Adults and adolescents. Developmental Psychology, 29 549-563. Birndorf, S., Ryan, S., Auinger, P., & Aten, M. (2005). High self-e steem among adolescents: Longitudinal trends, sex differen ces, and protective factors. Journal of Adolescent Health, 37 194-201. Bleakley, A., & Ellis, J. (2003). A role for public health research in shaping adolescent health policy. American Journal of Public Health, 93 1801-1802. Bogaert, A. & Fisher, W. (1995). Predictors of university mens number of sexual partners. The Journal of Sex Research, 32 119-130. Brady, S., & Donenberg, G. (2006). Mechanisms linking violence exposure to health risk behavior in adolescence: Motivati on to cope and sensation seeking. Journal of the American Academy of Child and Adolescent Psychiatry, 45 673-680. Brewer, B. & Cornelius, A. (2001). Norms and factorial invariance of th e Athletic Identity Measurement Scale. Academic Athletic Journal, 15 103-113. Brewer, B., Van Raalte, J., & Linder, D. ( 1993). Athletic identity: Hercules muscles or Achilles heel? International Journal of Sport Psychology, 24 237-254. Brindis, C. (2006). A public health success: Understanding policy changes related to teen sexual activity and pregnancy. American Review of Public Health, 27 277-295.
131 Bronfenbrenner, U. (2000). Ecological systems theory. In A. Kazdin (Ed.), Encyclopedia of psychology Washington DC, and New York: Amer ican Psychological Association and Oxford University Press. Bronfenbrenner, U., & Morris, P. (1998). The ecology of developmental processes. In W. Damon (Ed.), Handbook of child psychology (5th ed., Vol. 1). New York, NY:Wiley. Brook, J., Adams, R., Balka, E., Whiteman, M., Zhang, C., & Sugerman, R. (2004). Illicit drug use and risky sexual behavior among Afri can American and Puerto Rican urban adolescents: The longitudinal links. The Journal of Genetic Psychology, 165 203220. Brown, C., & Hartley, D. (1998). Athletic identity and career maturity of male college student athletes. International Journal of Sport Psychology, 29 17-26. Bruckner, H., & Bearman, P. (2005). After the promise: The STD consequences of adolescent virginity pledges. Journal of Adolescent Health, 36 271-278. Bryman, A. (2004). Social Research Methods, Second Edition. Oxford: Oxford University Press. Buysse, A., & Van Oost, P. (1997). Appropriate male and female safer sexual behaviour in heterosexual relationships. AIDS Care, 9 549-561. Corcoran, J. (1999). Ecological factors associated with adolescen t pregnancy: A review of the literature. Adolescence, 34 603-619. Crissy, S., & Honea, J. (2006). The relationship between athletic participation and perceptions of body size and weight control in adoles cent girls: The role of sport type. Sociology of Sport Journal, 23 248-272. Crockett, L., Bingham, C., Chopak, J., & Vicary, J. (1996). Timing of firs t sexual intercourse: The role of social control, soci al learning, and problem behavior. Journal of Youth and Adolescence, 25 89-111. Crothers, L., Field, J., & Kolber t, J. (2005). Navigating power control, and being nice: Aggression in adolescent girls friendships. Journal of Counseling and Development, 83 349-354. de Vaus, D. (2004). Research Design in So cial Research. London: SAGE Publications. Dillman, D. (2000). Mail and Intern et Surveys. New York, NY: Wiley. Dodge, T. & Jaccard, J. (2002). Participation in athletics and fe male sexual risk behavior: The evaluation of four causal structures. Journal of Adolescent Research, 17 42-67.
132 Donohew, L., Zimmerman, R., Cupp, P., Novak, S., Colon, S., & Abell, R. (2000). Sensation seeking, impulsive decision-making, and risky sex: Implications for risk-taking and design of interventions. Personality and Individual Differences, 28 1079-1091. Douglas, K. (2002). The impact of schools and school programs upon adolescent sexual behavior. Journal of Sex Research, 39 27-33. Egloff, B., & Gruhn, A. (1996). Personality and endurance sports. Personality and Individual Differences, 21 223-229. Feldman, S., Turner, R., & Araujo, K. (1999). In terpersonal context as an influence on sexual timetables of youths: Gender and ethnic effects. Journal of Research on Adolescence, 9 25-52. Frankenberger, K. (2004). Adol escent egocentrism, risk per ceptions, and sensation seeking among smoking and nonsmoking youth. Journal of Adolescent Research, 19 576-590. Etzkin, R. (2004). How parenting style and religio sity affect the timing of Jewish adolescents sexual debut [electronic resource]. Masters thesis. Gainesville, FL; University of Florida 2004. Greene, K., Krcmar, M., Walters, L., Rubin, D., & Hale, J. (2000). Targeting adolescent risktaking behaviors: The c ontributions of egocentris m and sensation-seeking. Journal of Adolescence, 23 439-461. Grove, J., Fish, M., & Eklund, R. (2004). Changes in athletic identity following team selection: Self-protection versus self-enhancement. Journal of Applied Sport Psychology, 16 7581. Gruber, E. & Machamer, A. (2000). Risk of school failure as an early indicator of other health risk behaviour in American high school students. Health, Risk & Society, 2 59-68. Grunbaum, J., Kann, L., Kinchen, S., Ross, J., Hawkins, J., Lowry, R., et al. (2004). Youth risk behavior surveillance Un ited States, 2003 (abridged). Journal of School Health, 74 307-324. Guest, A. & Schneider, B. (2003). Adolescents extracurricular particip ation in context: The mediating effects of schools, communities, and identity. Sociology of Education, 76 89-109. Gupta, R., Derevensky, J., & Ellenbogen, S. (2006). Personality characteri stics and risk-taking tendencies among adolescent gamblers. Canadian Journal of Behavioural Science, 38 201-213.
133 Halpern, C., Hallfors, D., Bauer, D., Iritani, B., Waller, M., & Hyunsan, C. (2004). Implications of racial and gender differences in patterns of adolescent ri sk behavior for HIV and other sexually transmitted diseases. Perspectives on Sexual & Reproductive Health, 36 239247. Hartman, M., & Rawson, H. (1992). Differences in and correlates of sensation seeking in male and female athletes and nonathletes. Personality and Individual Differences, 13 805812. Herman-Stahl, M., Krebs, C., Kroutil, L., & Heller D. (2006). Risk and protective factors for nonmedical use of prescription stimulan ts and methamphetamine among adolescents. Journal of Adolescent Health, 39 374-380. Horton, R., & Mack, D. (2000). Athletic ident ity in marathon runners: Functional focus or dysfunctional commitment? Journal of Sport Behavior, 23 101-119. Hoyle, R., Fejfar, M., & Miller J. (2000). Personality and sexual risk taking: A quantitative review. Journal of Personality, 68 1203-1231. Hoyle, R., Stephenson, M., Palmgreen, P., Lorch, E., & Donohew, R. (2002). Reliability and validity of a brief measur e of sensation seeking. Personality and Individual Differences, 32 401-414. Huba, G., Newcomb, M., & Bentler, P. (1981) Comparison of canoni cal correlation and interbattery factor analysis on sens ation seeking and drug use domains. Applied Psychological Measurement, 5 291-306. Husler, G., & Plancheral, B. (2006). A gender specific model of substance use. Addiction Research & Theory, 14 399-412. Jack, S. & Ronan, K. (1998). Sensation seeking among highand lowrisk sports participants. Personality and Individual Differences, 25 1063-1083. Jackson, T. (2006). Effects of sibling relationships and religios ity on the timing of sexual debut [electronic resource]. Masters thesis. Gainesville, FL; University of Florida 2006. Jacobsen, K., & Crockett, L. (2000). Parent al monitoring and adolescent adjustment: An ecological perspective. Journal of Research on Adolescence, 10 65-97. Kim, J., & Mueller, C. (1978). Factor Analysis: Statistical Methods and Practical Issues. Beverly Hills, CA: Sage Publications. Kirkcaldy, B., Shephard, R., Siefen, R. (2002). The relationship between physical activity and self-image and problem be haviour among adolescents. Social Psychiatry and Psychiatric Epidemiology, 37 544-550.
134 La Greca, A., Prinstein, M., & Fetter, M. (2001) Adolescent peer crowd affiliation: Linkages with health-risk behavior s and close friendships. Journal of Pediatric Psychology, 26 131-143. Lamb, S. (1997). Sex education as moral education: Teaching fo r pleasure, about fantasy, and against abuse. Journal of Moral Education, 26 301-315. Langer, L., Warheit, G., & McDonald, L. (2001) Correlates and predictors of risky sexual practices among a multi-racial/ethnic sample of university students. Social Behavior and Personality, 29 133-144. Lantz, C., & Schroeder, P. (1999). Endorsement of masculin e and feminine gender roles: Differences between participation in and identification with an athletic role. Journal of Sport Behavior, 22 545-557. Lau, P., Fox, K., & Cheung, M. (2005). Psychos ocial and socio-environm ental correlates of sport identity and sport participa tion in secondary school-age children. European Journal of Sport Science, 4 1-21. LEngle, K., Brown, J., & Kenneavy, K. (2006). The mass media are an important context for adolescents sexual behavior. Journal of Adolescent Health, 38 186-192. Lehman, S. & Koerner, S. (2004). Adoles cent womens sports involvement and sexual behavior/health: A proce ss-level investigation. Journal of Youth and Adolescence, 33 443-455. Little, C. & Rankin, A. (2001). Why do they st art it? Explaining reported early-teen sexual activity. Sociological Forum, 16 703-729. Longmore, M., Manning, W., Giordano, P., & Rudol ph, J. (2004). Self-esteem, depressive symptoms, and adolescents sexual onset. Social Psychology Quarterly, 67 279-295. Mancini, J. & Huebner, A. ( 2004). Adolescent risk behavior pa tterns: Effects of structured time-use, interpersonal connections, self-sys tem characteristics, and socio-demographic influences. Child and Adolescent Social Work, 21 647-668. Mandara, J., Murray, C., & Bangi, A. (2003). Predic tors of African American adolescent sexual activity: An ecological framework. Journal of Black Psychology, 29 337-356. Martens, M., Watson, J., & Beck, N. (2006). Sport-type differences in alcohol use among intercollegiate athletes. Journal of Applied Sport Psychology, 18 136-150. Maselko, J., & Kubzansky, L. (2006). Gender di fferences in religious practices, spiritual experiences and health: Results from the U.S. general social survey. Social Science and Medicine, 62 2848-2860.
135 Meschke, L., Zweig, J., Barber, B., & Eccles J. (2000). Demogr aphic, biological, psychological, and social predictors of the timing of first intercourse. Journal of Research on Adolescence, 10 315-338. Michels, T., Kropp, R., Eyre, S., & Halpern-Felsher, B. (2005). Initia ting sexual experiences: How do young adolescents make decisions regarding early sexual activity? Journal of Research on Adolescence, 15 583-607. Miller, B. (2002). Family influences on a dolescent sexual and contraceptive behavior. Journal of Sex Research, 39 22-26. Miller, K., Farrell, M., Barnes, G., Melnick, M., & Sabo, D. (2005). Gender/racial differences in jock identity, dating, and adolescent sexual risk. Journal of Youth and Adolescence, 34 123-136. Miller, J., Lynam, D., Zimmerman, R., Logan, T ., Leukefeld, C., & Clayton, R. (2004). The utility of the Five Factor Model in understanding risky sexual behavior. Personality and Individual Differences, 36 1611-1626. Miller, K., Sabo, D., Farrell, M., Barnes, G., & Melnick, M. (1998). Athletic participation and sexual behavior in adolescents: The di fferent worlds of boys and girls. Journal of Health and Social Behavior, 39 108-123. Moore, S., & Rosenthal, D. (1991). Adolescents perceptions of friends and parents attitudes to sex and sexual risk-taking. Journal of Community and Applied Social Psychology, 1 189-200. Nixon, H. (1997). Gender, sport, and a ggressive behavior outside sport. Journal of Sport and Social Issues, 21 379-391. ONeil, J. (1995). Patterns of gender role conflict and strain: Sexism and fear of femininity in mens lives. The Personnel and Guidance Journal, 60 203-210. OSullivan, D., Zuckerman, M., & Kraft, M. (1998). Personality characteristics of male and female participants in team sports. Personality and Individual Differences, 25 119-128. Palapattu, A., Kingery, J., & Ginsburg, G. ( 2006). Gender role orie ntation and anxiety symptoms among African Am erican adolescents. Journal of Abnormal Child Psychology, 34 423-431. Palmgreen, P., Donohew, L., Lorch, E., Hoyle, R., & Stephenson, M. (2001). Television campaigns and adolescent marijuana use: Tests of sensation seeking targeting. American Journal of Public Health, 76 525-531.
136 Perkins, D., Luster, T., Villarru el, F., & Small, S. (1998). An ecological, risk-factor examination of adolescents sexual activity in thr ee ethnic groups. Journal of Marriage and the Family, 60 660-673. Petipas, A., Van Raalte, J., Cornelius, A., & Pres brey, J. (2004). A life skills development program for high school student-athletes. The Journal of Primary Prevention, 24 325334. Pyle, R., McQuivey, R., Brasington, G., & Steine r, H. (2003). High school student athletes: Associations between intensity of participation and health factors. Clinical Pediatrics, 42 697-702. Richards, M., Miller, B., ODonnell, P., Wasserm an, M., & Colder, C. (2004). Parental monitoring mediates the effects of age and sex on problem behaviors among African American urban young adolescents. Journal of Youth and Adolescence, 33 221-233. Rodgers, K. (1999). Parenting processes related to sexual risk-taking be haviors of adolescent males and females. Journal of Marriage and the Family, 61 99-109. Rosenthal, D., Smith, A., & de Visser, R. (1999). Personal and social fact ors influencing age at first sexual intercourse. Archives of Sexual Behavior, 28 319-333. Rostosky, S., Wilcox, B., Comer-Wright, M., & Randall, B. (2004). The impact of religiosity on adolescent sexual behavior: A review of the evidence. Journal of Adolescent Research, 19 677-697. Rudd, A. & Stoll, S. (2004). What type of character do athletes possess? An empirical examination of college athletes versus college nonathletes with the RSBH Value Judgment Inventory. Sport Journal, 7 Retrieved April 10, 2007, from http://www.the sportjournal.org/2004Journal /Vol7-No2/RuddStoll.asp. Ryska, T. (2002). The effects of athletic id entity and motivation goals on global competence perceptions of student-athletes. Child Study Journal, 32 109-130. Santrock, J. (2001). Adolescence (5th ed.) New York, NY: McGraw-Hill. Savage, M. & Holcomb, D. (1999) Adolescent female athletes sexual risk-taking behaviors. Journal of Youth and Adolescence, 28 595-603. Schaalma, H., Abraham, C., Gillmore, M., & K ok, G. (2004). Sex education as health promotion: What does it take? Archives of Sexual Behavior, 33 259-269. Schroth, M. (1995). A comparis on of sensation seeking among diffe rent groups of athletes and nonathletes. Personality and Individual Differences, 18 219-222.
137 Schuster, M., Bell, R., & Kanouse, D. (1996). The sexual practices of adolescent virgins: Genital sexual activities of high school students who have ne ver had vaginal intercourse. American Journal of Public Health, 86 1570-1576. Scourfield, J., Stevens, D., & Merikangas, K. (1996). Substance abuse, comorbidity, and sensation seeking: Gender differences. Comprehensive Psychiatry, 37 384-392. Silva, J. (1983). The perceived legitimacy of rule violating behavior in sport. Journal of Sport Psychology, 5 438-448. Small, S. & Luster, T. (1994). Adolescent sexua l activity: An ecological, risk-factor approach. Journal of Marriage and the Family, 56 181-192. Smith, E. & Caldwell, L. (1994). Participati on in high school sports and adolescent sexual activity. Pediatric Exercise Science, 6 69-74. Smith, C., Denton, M., Faris, R., & Regnerus, M. (2002). Mapping American adolescent religious participation. Journal for the Scientif ic Study of Religion, 41 597-612. Smith, L. & Guthrie, B. (2005). Testing a m odel: A developmental perspective of adolescent male sexuality. Journal for Specialists in Pediatric Nursing, 10 124-138. Spitalnick, J., Diclemente, R., Wi ngood, G., Crosby, R., Milhausen, R., Sales, J, et al. (2007). Brief report: Sexual sensati on seeking and its relationship to risky sexual behavior among African-American adolescent females. Journal of Adolescence, 30 165-173. Stanton, B., Li, X., Cottrell, L., & Kaljee, L. (2001). Early in itiation of sex, drug-related risk behaviors, and sensation-seeking am ong urban, low-income African-American adolescents. Journal of the American Medical Association, 93 129-138. Stephenson, M., Palmgreen, P., Hoyle, R., Donohew L., Lorch, E., & Colon, S. (1999). Shortterm effects of an anti-marijuana medi a campaign targeting high sensation seeking adolescents. Journal of Applied Communication Research, 27 175-195. Straub, W. (1982). Sensation seeking am ong high and low-risk male athletes. Journal of Sport Psychology, 4 246-253. Suitor, J. & Reavis, R. (1995). Football, fast cars, and cheerleading: Adolescent gender norms, 1978-1989. Adolescence, 30 265-272. Tapert, S., Aarons, G., Sedlar, G., & Brown, S. (2001). Adoles cent substance use and sexual risk-taking behavior. Journal of Adolescent Health, 28 181-189. Tasiemski, T., Kennedy, P., Gardner, B., & Blaikley, R. (2004). Athletic identity and sports participation in people with spinal cord injury. Adapted Physical Activity Quarterly, 21 364-378.
138 Tayler-Seehafer, M. (200). Risky sexual behavior among adolescent women. Journal of the Society of Pediatric Nurses, 5 15-25. Tremblay, L. & Frigon, J.. (2004). Biobehavioura l and cognitive determin ants of adolescent girls involvement in sexual risk behaviour s: A test of three theoretical models. The Canadian Journal of Human Sexuality, 13 29-43. Upchurch, D., Levy-Storms, L., Sucoff, C., & Aneshensel, C. (1998). Gender and ethnic differences in the timing of first sexual in tercourse. Family Planning Perspectives, 30 121-127. van Beurden, E., Zask, A., Brooks, L., & Dight, R. (2005). Heavy episodic drinking and sensation seeking in adolescence as predic tors of harmful driving and celebrating behaviors: Implications for prevention. Journal of Adolescent Health, 37 37-43. Walker, S. (1997). When no becomes yes: Why girls and women consent to unwanted sex. Applied and Preventive Psychology, 6 157-166. Ward, L. & Friedman, K. (2006). Using TV as a guide: Associations between television viewing and adolescents se xual attitudes and behavior. Journal of Research on Adolescence, 16 133-156. Williams, R., & Ricciardelli, L. (2003). Negative perceptions about self-control and identification with gender-role stereotypes related to binge eating, problem drinking, and to co-morbidity among adolescents. Journal of Adolescent Health, 32 66-72. Young, S. (2002). Substance use, abuse, and de pendence in adolescence: Prevalence, symptom profiles and correlates. Drug & Alcohol Dependence, 68 309-322. Zarevski, P., Marusic, I., Zolotic, S., Bunjevac, T., & Vukosav, Z. (1998). Contribution of Arnetts inventory of sensation seeking and Zuckermans sensation seeking scale to the differentiation of athletes engage d in high and low risk sports. Personality and Individual Differences, 25 763-768. Zuckerman, M. (1979) Sensation seeking: Beyond the optimal level of arousal. Hillsdale, NJ: Erlbaum. Zuckerman, M., & Link, K. (1968) Construct validity for the sensation-seeking scale. Journal of Consulting and Clinical Psychology, 32 420-426.
139 BIOGRAPHICAL SKETCH Sally Moore was born in Winter Park, Florida an d grew up in Tampa, Florida. She started her first gymnastics class when she was eight ye ars old and continued her career as a competitive gymnast until the age of sixteen, when an injury fo rced her to retire. She remained active in the sport by coaching young gymnasts. Sally gradua ted high school in 2005 from Lee Academy for Gifted Education and moved to Gainesville to atte nd the University of Florida. After earning her B.S. in Psychology in 2007, she entered the M.S. program in the department of Family, Youth and Community Sciences. She will begin work to ward her Ph.D. in School Psychology at UF in the fall. In addition to her acad emic endeavors, Sally has worked as a preschool teacher and intends to pursue an early childhood specializ ation as part of he r doctoral program.