|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 RELIGIOSITY-RISK BEHAVIOR MODEL By WENDI ANN MALONE 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 2008
2 2008 Wendi A. Malone
3 To my Aunt Rhonda
4 ACKNOWLEDGMENTS I thank Jam es Shepperd for his guidance and support. In addition, I would like to thank Kate Sweeny, Cathy Cottrell, James Algina, and Julia Graber for their invaluable comments and suggestions.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 LIST OF FIGURES.........................................................................................................................8 ABSTRACT.....................................................................................................................................9 CHAP TER 1 INTRODUCTION..................................................................................................................10 Adolescent Risk Behavior...................................................................................................... 10 Historical Background of the link be tween Religion and Risk Behavior ...............................11 Potential Mechanisms of the link between Religion and Risk Behavior................................ 13 Criticisms of Research on Religiosity and Risk Behavior...................................................... 15 Religiosity-Risk Behavior Model...........................................................................................16 Study Overview................................................................................................................. .....20 Hypotheses..............................................................................................................................20 2 METHOD......................................................................................................................... ......23 Participants.............................................................................................................................23 Procedure................................................................................................................................23 Materials.................................................................................................................................23 3 RESULTS...............................................................................................................................27 4 DISCUSSION.........................................................................................................................36 APPENDIX A CONSENT FORM.................................................................................................................. 42 B DEMOGRAPHIC QUESTIONNAIRE.................................................................................. 43 C RELIGIOUS COMMITMENT INVENTORY (RCI-10)...................................................... 45 D PEER/ADULT MODELS......................................................................................................46 E RELIGIOUS COPING STRATEGIES.................................................................................. 47 F SUBJECTIVE NORMS..........................................................................................................48
6 G PROTOTYPES.......................................................................................................................49 H RISK PERCEPTIONS............................................................................................................ 50 REFERENCES..............................................................................................................................51 BIOGRAPHICAL SKETCH.........................................................................................................55
7 LIST OF TABLES Table page 3-1 Correlations for model constructs...................................................................................... 303-2 Means and standard deviations for model constructs........................................................ 313-3 Frequencies for lifetime illegal drug use and drug use in the past month......................... 323-4 Summary statistics of the religiosity and risk behavior structural equation models.......... 33
8 LIST OF FIGURES Figure page 1-1 Religiosity and Risk Behavior Model................................................................................ 223-1 Model 1: The hypothesized model..................................................................................... 343-2 Model 4: The final model.................................................................................................. 35
9 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 RELIGIOSITY-RISK BEHAVIOR MODEL By Wendi A. Malone December 2008 Chair: James Shepperd Major: Psychology I propose a framework for understanding religios ity and risk behavior (Religiosity and Risk Behavior Model) that specif ies a variety of upstream features of religiosity (prescriptions for behavior, peer/adults models, decreased oppo rtunity, and coping strate gies) and downstream features of religiosity (risk related cognitions, ri sk perceptions, decreased behavioral willingness) that are responsible for lower rates of ri sky behavior among religious adolescents. Undergraduates (N = 272) comple ted an online questionnaire that measured religiosity, illicit drug use, and the upstream and downstream features of religiosity. Analys is using structural equation modeling generally supported our theore tical model. Religious adolescents were less likely to report using illicit drugs and the li nk between religiosity and drug use was mediated through paths specified in our model.
10 CHAPTER 1 INTRODUCTION Adolescent Risk Behavior Adolescence is a tim e of increased risky beha vior. A recent survey revealed that 19% of high school seniors reported smoki ng cigarettes on a regular basi s, 49.8% reported using alcohol in the past 30 days, and 29.7% binge drank in the past two weeks (Johnston, OMalley, & Bachman, 2002). Although some ri sk-taking may be a positive force in development (Baumrind, 1987; Moore & Parsons, 2000; Shedler & Block, 1990), chronic risk-taking and risky activities such as illegal drug use and unprotected sexual in tercourse can be dangerous. Risk behavior can produce long-term negative consequences. For instance, each year approximately one million adolescent girls become pregnant in the United States (Fielding & Williams, 1991), and 3 million adolescents are diagnosed with a sexually tran smitted disease (Centers for Disease Control & Preventions, 2004). Importantly, not all adolescents engage in risky behavior. For example, girls are less likely than boys to engage in risky behavior (Tuinstra, Groothoff, Van Den Heuvel, & Post, 1998; Weden & Zabin, 2005) and younger adolescent s are less likely than older adolescents (Resnick, Bearman, Blum, Bauman, Harris, Jones, et al., 1997). These demographic differences aside, one individual difference vari able consistently linked to lowe r risk behavior is religiosity. Adolescents who score high on measures of religios ity (religious adolescents) are less likely than adolescents who score low on such measures (non-religi ous adolescents) to e ngage in a host of risky behavior. For example, reli gious adolescents report that they are less likely to smoke cigarettes (Amey, Albrecht, & Miller, 1996; Dunn, 2005; Nonnemaker et al., 2003; Steinman & Zimmerman, 2004; Wallace & Forman, 1998), drink alcohol (Amey et al., 1996; Cochran & Akers, 1989; Dunn, 2005; Nonnemaker et al., 20 03; Steinman & Zimmerman, 2004; Wallace &
11 Forman, 1998), use illicit drugs (Adlaf & Smart, 1985; Dunn, 2005; Miller, Davies, & Greenwald, 2000; Steinman & Zimm erman), and engage in sexual intercourse (DuRant, et al., 1990; Nonnemaker et al., 2003). In addition, religious adolescents are less likely to engage in violent behavior (Nonnemaker et al., 2003; Powell, 1997), and a ttempt suicide (Nonnemaker et al., 2003). These studies demonstrate that religiosity is an important correlate of risk behavior. They are limited, however, in that they do not pr ovide insight into why religious adolescent are less likely to engage in risk behavior. The ai m of the current investigation is to move the exploration of the link between religiosity and risk behavior beyond simple correlations by proposing a theoretical model that attempts to capture and understa nd this relationship. I will test the theoretical model on an undergraduate student sample. Although the typical age range of undergraduates falls at late adolescence, it nevertheless provides useful information. Consequences of risk behavior (e.g., unplanned pregnancy, drug addictio n) can be just as destructive to late adolescents as to early and middle adolescence. Historical Background of the Link Betw een Religion and Risk Behavior Much of the research examining religiosity before the late 1960s was problematic in that it often lacked a comparison group and results appeared biased in favor of the investigators views on religion (Powers, 1967). One notable exception was a study by Hirschi and Stark (1969) that explored the relati onship between religiosity and delinquency. Hirschi and Stark found that church attendance was unrelated to six types of deli nquent behavior. These delinquent behaviors included stealing (items less than $2, items between $2 and $50, and items over $50), stealing a car for a joy ride, banging up property, and fighting. The authors concluded that adolescents who attend church regu larly are just as likely as a dolescents who rarely or never attend church to commit delinquent acts. Religion simply did not matter.
12 Although methodologically, th e study by Hirschi and Star k (1969) represented an improvement over prior studies of religiosity and risk behavi or, it nevertheless had several limitations. First, Hirschi and Stark only exam ined a subset of delinquent behaviors (e.g., vandalism, assault, and larceny). Second, their co nclusion that religion does not matter is too simplistic. Burkett and White (1974) argued that religion may do a good job of deterring delinquent behavior, but secula r institutions (e.g., law enforcement, schools) may also deter delinquency. The behaviors examined by Hirsch i and Stark include vandalism, assault, and larceny. Religious institutions do not condone behaviors such as stealing, but neither do school officials or the police. Thus, there are severa l potentials sources from which adolescents can learn that stealing is wrong. Relig ion may not be needed to deter adolescents from stealing. On the other hand, religion should predict individua l differences in engaging in victimless delinquent acts (e.g., drinking alco hol) because secular controls are less clear on these issues. Drinking alcohol under the age of 21 is illegal, bu t through the media, peers, and even parents, adolescents receive messages that condone underage drinking. Consistent w ith the prediction that religion is related to engagement in victimle ss delinquent acts, religi ous adolescents are less likely to drink alcohol and smoke marijuan a (Amey et al., 1996; Burkett & White, 1974; Cochran & Akers, 1989; Dunn, 2005; Nonnemaker et al., 2003; Steinman & Zimmerman, 2004; Wallace & Forman, 1998). Second, researchers have questioned the external validity of the Hirschi and Stark findings. Their sample consisted of middle and high school st udents in Northern Calif ornia. Researchers in other parts of the country such as Atlanta (Higgins and Albrecht, 1977), Utah (Albrecht, Chadwick, & Alcorn, 1977), Arizona (Jensen & Erickson, 1979), and Idaho (Albrecht et al.,
13 1977) used procedures similar to Hirschi and Stark yet found a negative relationship between religion and delinquency greater religiosity corresponded to lo wer delinquent behavior. More recently, Stark (1996) proposed that re ligion deters delinquency only in regions where the majority of people are actively re ligious. Thus, no relationship emerged between religion and delinquency in Northern California sa mple because of low levels of religiosity on the west coast of the United States. Rates of chur ch attendance in this region of the country are low (e.g., the rate of church atte ndance in Seattle is the lowest of any metropolitan area at 280 per 1,000), compared to others re gions where the predicted relationship was found (e.g., the rate of church attendance in Provo, Utah is 966 per 1,000). In support of this notion, Stark (1996) found th at the correlation between church attendance and getting in trouble with the law was highest in the East (-.32), Midwest (-.36), and South (.39), where church attendance is roughly 60%, lo wer in the Mountain region (-.23) where church attendance is roughly 48%, and ne ar zero (-.02) in the Pacific region where church attendance is roughly 36%. Although this finding is supportive, the author acknowle dges that region is a crude measure of the religious sentiments of a communit y. Further, this type of explanatory framework ignores several factors of an i ndividual and his/her environment (e.g., attitudes, risk perceptions) that likely influence willingness to engage in risk behavior. Potential Mechanisms of the Link betw een Religion and Risk Behavior Although several studies find a link between re ligiosity and risky be havior, researchers have a poor understanding of why the link exists. Little is known about what qualities religion offers that dissuade adolescents from engaging in risky behavior. Only a handful of studies have examined the mechanisms by which religion even tuates in lower risk behavior. Two possible mechanism are action-specific beliefs and peer associations. An action-specific belief with regard to religion is a belief th at certain behaviors are sinful (e.g., using marijuana is sinful
14 behavior). Action-specific events repres ent a link between reli gious commitment and delinquency. Peer associations refe r to associations with people w ho engage in the risky behavior (e.g., adolescents who smoke mariju ana). Presumably, religious adol escents are inclined to adopt action-specific beliefs and the action-specific be liefs deter relevant ri sky behavior. Likewise, religious adolescents are inclined to avoid peers who smoke mariju ana and thus are less inclined to smoke marijuana themselves. Consistent with this reasoning was a study that assessed religious commitment (a measure of church attendance and religious identification), action-specific beliefs regarding marijuana, peer association with others who use marijuan a, and marijuana use (B urkett & Warren, 1987). Results indicated that religious commitment was an important pred ictor of marijuana use, but the effect of religious commitment was mediat ed by the action-specific belief that smoking marijuana is a sin and by peer associations. Reli gious adolescents appear to select peers who are similar in attitudes and behavior regarding the use of marijuana. This peer group provides an effective moral community which supports beliefs which inhibit th e use of marijuana (p. 127). A third mechanism that may dissuade religious adolescents from engaging in risk behavior is respect for the juvenile court system. Respect for the juve nile court system reflects how adolescents feel, and how they believe their fr iends feel toward police, probation officers, juvenile court judges, and the juvenile cour t. One study examined whether respect for the juvenile court system possibly mediated the re lationship between church attendance and 17 types of delinquent behavior (e.g., dr inking alcohol, stealing, skippi ng school; Higgins & Albrecht, 1977). Church attendance was negatively related to delinquent behavior and positively related to self and friends respect for the juvenile court sy stem. Further, self and friends respect for the juvenile court system was negatively related to self-reported delinquent behavior. Although the
15 authors did not directly test for mediation, they concluded from the pattern of correlations that respect for the juvenile court system is a possibl e mediator of the relationship between church attendance and delinquency. Criticisms of Research on Re ligiosity and Risk Behavior Research ex amining the relationship between religiosity and risk behavior has three limitations that have constrained progress in the area. First, much of the research on religiosity lacks theoretical basis. In many instances it a ppears that researchers tacked on a measure of religiosity at the end of a survey for explorat ory purposes without givi ng much thought to what they might find or why religiosity would be relate d to risk behavior. For instance, one review of 65 published studies found that two-thirds were a-theoretical (Tittle & Welch, 1983). A more recent review revealed that over half of the studies published between 1998 and 2003 that examined religion, spirituality, and health behavior lacked th eoretical grounding (Rew & Wong, 2006). As noted earlier, some researchers have implied theory by exploring mechanisms (e.g., respect for the juvenile court systems) that may mediate the link between church attendance or other features of religiosity and decreased risk behavior. However, lacking is any overarching theoretical model that integrates previous fi ndings and describes the various paths by which religiosity leads to lower risk behavior. Sec ond, few studies examine the relationship between religiosity and risky behavior longitudinally. Most studies ar e cross-sectional, examining religiosity and risk behavior simultaneously, making it impossible to examine how religiosity measured at one point in time is related to risk behavior at a later point in time. Third, there is no consensus in the prior studies of religiosity and risk behavior in how to define religiosity. The lack of clarity in defining religiosity has led to inconsistency in measurement. Some researchers have operationali zed religiosity in terms of affiliation with a religious organization (Takyi, 2003) Others have operationalized re ligiosity in behavioral terms
16 such as frequency of attending religious services (Lorch & Hughes, 1985; Mullen & Francis, 1995). These two operations are limited in that they define adolescents who do not attend religious services as non-religious even though th ey may have strong religious sentiments. To address this limitation, other resear chers have operationalized reli giosity as a self-judgment (e.g., How religious are you?). Complicating matters, many researchers re ly on single-item measures, which are problematic because of their questionable reliability. In addition, it is unclear to what extent the various approaches to assessing religiosity (re ligious affiliation, religious event attendance, frequency of prayer, religious sentiments) are correlated and thus ta pping the same underlying construct. It is noteworthy that some research has assessed religiosit y with reliable, valid measures (e.g., Cornwall, Albrecht, Cunningham, & Pitcher, 1986). However, the scale items assess Christian religiosity and exclude other religious faiths (Hill & Hood, 1999). Moreover, the researchers have not used their inst ruments to predict risk behavior. Religiosity-Risk Behavior Model To address the limitation that previous rese arch on religiosity l acks theoretical grounding, I developed The Religiosity-Risk Behavior Mode l (see Figure 1). This model incorporates theoretical links identified by ot her researchers (e.g., action-specific beliefs, peer associations). In addition, the model proposes that a religious faith offers several f eatures to religious adolescents that dimini sh risky behavior. The upstream features include prescriptions for behavior, access to peer and adult models, decreased risk opportunity, and positive coping strategies. The downstream features include cognitions, risk pe rceptions, and behavioral willingness. Upstream features of religiosity Several upstream features of religious faiths may contribute to decreased risky behavi or. First, all religious faiths have prescriptions for behaviors
17 and these prescriptions are often at odds with engagement in risky behavior. The prescriptions provide specifications of what behavior is or is not acceptable or moral. The prescriptions may be explicit (e.g., Thou shall not steal.) or imp lied (e.g., Honor thy father and thy mother can imply quite a few things about personal behavi or). Second, religion provides access to a network of peers and adults who, among other things, model non-risky behavior and care about and monitor the behavior and well being of younger me mbers. This network of peers and adults recognizes the important role identified in prior research on religiosity an d risk behavior that peers and others play in the decision to engage in risky behavior (Burkett & Warren, 1987). Third, adolescents engaged religious activitie s often have less opport unity than secular adolescents to engage in risk behavior. Fourth, certain risk be haviors (e.g., consuming alcohol and using drugs) represent attempts to cope w ith negative and stressful life events. Religious faiths sometimes provide alternative strategies for coping with life stressors (Shafranske, 1992). These alternative coping strategies may include praying for guidance, asking others to pray for them, and seeking clergy for counsel or assistance. As evident in Figure 1, three of the upstream features of religious faiths (behavioral prescriptions, decreased opportuni ty and coping strategies) can ha ve a direct effect on risky behavior. To the extent that adolescents accept a prescription for behavior, they should be less likely to display risky behavior encompassed by the prescription. Likewise, to the extent that adolescents have less opportunity to engage in risky behavior, they are less likely to engage in risky behavior. Finally, the coping strategies provide adolescents with a means of dealing with stressful situations that represent alte rnatives to engaging in risky behavior. Downstream features of religiosity. Several of the upstream features of religious faiths affect engagement in risk behavi or indirectly through intermediate routes. First, a consequence of
18 decreased opportunity is the sust ained perception that engaging in risk behavior inevitably leads to negative outcomes. Specificall y, people base risk judgments in part on their experience with the behavior (driving while drunk) and their experience with the outcome (getting in an automobile accident). Typically there is not a one-to-one correspondence between the risk behavior and the outcome. In mo st instances, people who drive drunk do not get in automobile accidents. The experience of the behavior without the negative outcome serves to undermine the perception that the outcome is an inevitable cons equence of the behavior. The result is that people who engage in risky beha vior without negative conseque nces come to perceive the behavior as less risky, whereas people who do not engage in the risky behavior continue to perceive the behavior as risky. C onsistent with this reasoning is th e finding that adolescents with no experience with a risky behavi or (e.g., no experience drinking alc ohol) perceived their risk of experiencing a negative conse quence following the behavior (e .g., getting sick from drinking several beers) as greater than did adolescents who had experience with the behavior but no experience with the negative c onsequences (Halpern-Felsher et al, 2001). The (sustained) perception that risky behavior l eads to negative outcomes makes the behaviorally inexperienced adolescents unwilling to engage in risky behavior. Peer/adult models and prescriptions for beha vior can also indir ectly influence risky behavior. Peer/adult models can provide informa tion about the prescriptions for behavior. More importantly, both of these upstream features can influence risk related cognitions by affecting how adolescents think about risk behavior. Thes e risk cognitions incl ude attitudes, actionspecific beliefs, subjective norms and risk prototypes. Attitudes represent peoples likes or dislikes for an attitude object Peer/adult models as well as th e prescriptions for behaviors can prompt adolescents to form negative attitude s about risk behaviors. As noted earlier, Action
19 Specific Beliefs represent beliefs about behaviors (e.g., s moking marijuana is sinful; Burkett & Warren, 1987), and adolescents who endorse negativ e action-specific beli ef about negative or risky behavior should be less likely to engage in the corresponding ris ky behavior. Importantly, action specific beliefs are distinct from prescrip tions for behaviors. Prescriptions for behavior represent the extent to which participants have b een taught that a particular behavior is right or wrong. As such, prescriptions represent dos and donts Action specific beliefs represent thoughts about the acceptability particular actions. In a sense, they represent the extent to which participants endorse the prescriptions. Regarding subjective norms, according to the theory of r easoned action, people consider whether important others will approve or disappr ove of their behavior when forming behavioral intentions to engage in beha vior (Ajzen & Fishbein, 1980). Th e prescriptions for behaviors typically involve norms for acceptable behavior, and peers and adults both model and enforce compliance with the norms. To the extent that the norms against risky behavior are clear and accepted by the adolescents or enforced by peers and adults, then they should eventuate in less willingness to engage in risky behavior. Regarding prototypes people often have an image or pr ototype of the typical person who engages in a risky behavior. For example, most adolescents have a pr ototype of the typical person who uses drugs. Further, adolescents underst and that if they engage in risky behavior, they may be seen has having characteristics associ ated with the risk prototype (Gibbons, Gerrard, Blanton, & Russell, 1998). To the extent that adol escents perceive a prototype that engages in a risky behavior as undesirable, they will be di sinclined to engage in the risky behavior. Both risk cognitions and ri sk perceptions influence behavioral willingness Adolescents often do not intend to engage in risky behavior, but find themse lves in situations where the
20 opportunity to engage in risky beha vior is present. Adolescents differ in whether they are willing to engage in risky behavior when faced with the opportunity (Gibbons et al., 1998). Adolescents display greater willingness to engage in risky be havior when they perceive that norms support engaging in the behavior (Gibbons, Helweg-L arsen, & Gerrard, 1995) they have positive attitudes toward the behavior (Gibbons, Gerrard, Ouelette, & Bur zette, 1998), and their prototype of who engages in the behavior is positive (G ibbons et al., 1998). I propose that religious adolescents will display lower behavioral willingn ess because they are more likely to perceive that important others do not enga ge in, or approve of engaging in risky behaviors. I further propose that religious adolescents will hold more negative attitudes a bout engaging in risky behavior and more negative pe rceptions of the prototypical person who engages in risky behavior. Finally, I anticipate that the Religiosity-Risk Be havior Model will predict a variety of risky behaviors including drinki ng alcohol, smoking tobacco, usi ng illicit drugs, and engaging in sex (and engaging in sex without a condom). Study Overview Given past research, I expect that greater religiosity will correspond to lower risk behavior. Specifically, greater religiosity will correspond to lo wer levels drug use among our sample of late adolescents. For the current study, only illicit drug use was examined as an outcome variable. Although I anticipate that the model will predicts other risk behaviors described equally well, I specifically focused on illicit drug use. Hypotheses The prim ary aim is to examine why religious adolescents ar e less inclined than nonreligious adolescents to engage in illicit dr ug use. This primary aim will be accomplished by examining the various hypothesized links outlin ed in the model. Although the hypotheses are
21 phrased as though religiosity will be treated as a dichotomous pred ictor, all statistical analyses will treat religiosity as a continuous predictor. I predict the following. 1. Religious adolescents will be more likely th an non-religious adolescents to report having prescriptions against using illicit drugs. I fu rther predict that the prescriptions regarding illicit drug use will partially mediate the relationship between religiosity and drug. 2. Religious adolescents will be more likely th an non-religious adolescents to report having positive adults/peer models (i.e., adults and peers who care about the participants, and discourage or condemn drug use). I further predict that reports of peer/adult models will partially mediate the relationship between relig iosity and the prescriptions regarding drug use. 3. Prescriptions regarding illicit drug use and the availability of peer/adult models will be linked to participants risk cognitions (attitudes, norms, beliefs, and prototypes). Specifically, having prescriptions against dr ug use and positive peer/adult models will correspond with more negative attitudes toward using drugs, perc eiving drug use as nonnormative, holding more negative beliefs about using drugs, and more negative prototypes about the type of people who use drugs. 4. Religious adolescents, compared with non-religious adolescents, will have less favorable cognitions regarding drug use. Moreover, prescriptions against drug use and the availability of peer/adult mode ls will mediate the relationshi p between religiosity and the risk cognitions. 5. Religious adolescents will report having less opportunity to use drugs than will nonreligious adolescents. The less opportunity wi ll in turn correspond with greater risk attached to using drugs. Specifically, decrea sed opportunity will correspond with the belief that using drugs will result in addiction and punishment by law enforcement. Lower reported opportunity to use drugs wi ll correspond to lower drug use. 6. Negative cognitions about using drugs and a strong perception that using drugs will produce negative consequences will correspond to a lower willingness to use drugs. The lower willingness to use drugs will correspond to less drug use. Lower willingness to use drugs will mediate the relationship between the cognitions about using drugs and the perceived risk of using drugs. 7. Religious adolescents, compared with non-religi ous adolescents, will have more religious coping strategies for dealing with stressful s ituations. The availabili ty of religious coping strategies will be associated w ith a lower likelihood of turnin g to risky behavior such as using illicit drugs to deal with stressful situations. As a result the availability of religious coping strategies will partially mediate the re lationship between reli giosity and illicit drug use.
22 Figure 1-1. Religiosity and risk behavior model Opportunities Peer/Adult Models Cognitions: Attitudes, Norms, Beliefs, & Prototypes Religiosity Religious Coping Prescriptions for Behavior Risk Behavior (Drug Use) Upstream Features of Religiosity Downstream Features of Religiosity Risk Perceptions Behavioral Willingness
23 CHAPTER 2 METHOD Participants Participants were undergraduate students (190 male, 84 female) from psychology classes who completed an online survey of religiosity and risky behavior as part of a course obligation. Participants were predominantly Caucasian (Caucasian = 135, African American = 49, Asian American = 34, Hispanic = 43, Other = 13), betw een the ages of 18 and 21 (18 = 65, 19 = 90, 20 = 52, 21 = 41, 22 and over = 25, not reported = 1), a nd Protestant (Protest ant = 110, Catholic = 68, Agnostic = 26, Atheist = 10, Hindu = 6, Islami c = 6, Jewish = 16, Other = 13, not reported = 19). Procedure Participan ts who agreed to participate were di rected to a website th at contains all study materials. Once they consented to participate, participants were transferred to a webpage that contained the survey items. On completing the su rvey participants read a debriefing statement that described the purpose of the study and then were thank for their time. Materials Demographic items. After com pleting the consent fo rm, participants responded to several demographic items that assessed their age, sex (see Appendix B). Religious Commitment Inventory (RCI-10). The RCI-10 is a ten-item measure of the degree to which a person adheres to his or her religious values, beliefs, and practices and uses them in daily living (Worthington, et al., 2003, p.85). Example items include, I often read books and magazines about my faith. and My relig ious beliefs lie behind my whole approach to life. Participants will rate items on a scale from 1 = not at all true of me to 5 = totally true of me The RCI has acceptable psychometric charac teristics. For example, the scale has high
24 internal consistency (Cronbachs alphas above .8 5), test-retest reliabili ty over five months ( r = .84), and clear construct validity (W orthington, et al., 2003). (Appendix C) Prescriptions for behavior Prescriptions provide guide lines that determine what behaviors are or are not acceptable and moral. To measure this cons truct, participants completed an item that assessed the extent to which they have been exposed to specif ic prescriptions against illicit drug use. The item read as follows, I ha ve been taught that using drugs is wrong. (1 = strongly disagree to 7 = strongly agree). Peer/adults models Religion provides access to peers and adults who model non-risky behavior and care about and monitor the behavior and well being of religious adolescents. We used items such as, I know adults who care abou t me, I have peers who care about me, and The adults I know discourage or disapprove of risky behavior. For all items, participants responded with 1 = strongly disagree to 7 = strongly agree. (Appendix D) Opportunity Adolescents who are active in their religious institution may have less opportunity than secular adolescent s to engage in risk behavior. Participants perception of their opportunity to use illicit drugs wa s measured using the following ite m, It would be easy for me to get drugs Participants responded with 1 = strongly disagree to 7 = strongly agree Religious coping strategies We used five items from the Religious Coping Activities Scale (Pargament et al., 1990). Participants were di rected to think about a past stressful event and then to indicate, for example whether they, Found the lesson from God in the event, and Sought support from clergy. Par ticipants responded with 1 = not at all to 7 = a great deal. The subscales from which these items are dr awn have shown adequate reliability ( = .78-.92; Pargament, et al., 1990). (Appendix E)
25 Attitudes Attitudes represent peoples likes or dislik es for an attitude object. Participants rated their attitudes about drug use by completi ng the following item, Please rate how you feel about you using drugs. Items were rated on a scale from 1 = Very Negative to 7 = Very Positive Subjective norms. To measure subjective norms, I used items adapted from previous research (Gibbons, Gerrard, Blanton, & Russell, 1998) in which participants estimate of how often their friends and peers engage in various risk behaviors and how th eir friends and parents would react if the participant e ngaged in the risk behaviors. An example item is as follows, How many people your age do drugs or have used drugs? (Appendix F) Action-specific beliefs Action-Specific Beliefs represen t beliefs about behaviors and the acceptability of those behaviors. I used an it em adapted from previous research (Burkett & Warren, 1987; Burkett & White, 1974). Participants ra ted the following item, I believe using drugs is wrong on a scale from 1 = strongly disagree to 7 = strongly agree. Prototypes A prototype is a mental representation of the typical person who engages in a behavior. Participants read a description that defines the te rm prototype (Gibbons, Gerrard, & Boney McCoy, 1995). Then they rated prototyp es who use illicit drugs using the following adjectives: smart, confused, popular, immatu re, "cool" [sophisticated], self-confident, independent, careless, unattractive, dull [bori ng], considerate and se lf-centered (Gibbons, Gerrard, Blanton, & Russell, 1998). Participan ts responded to the adjectives with 1 = not at all to 7 = very. Factor analysis from previous research indicates that the tw elve adjectives load on three correlated factors labeled ma ture, self-assured, and attract ive (Blanton, Gibbons, Gerrard, Conger, & Smith, 1997; Gibbons et al., 1998). (Appendix G) Risk perceptions I used items modeled after those us ed in previous research exploring the extent to which people perceive negative out comes an inevitable consequence of engaging in
26 a given risky behavior (Halpern-Felsher et al, 2001). For exam ple, one of the drug use risk perception items asked participants to imagine they are at a party where they are offered the opportunity to try an illicit drug. They then estimated (from 0% to 100%) the likel ihood that they would overdose on the drug. See Appendix H. Behavioral willingness. We assessed behavioral willingness using items adapted from prior research (Gibbons et al ., 1998). Participants complete d the following item, Suppose you were with some friends at a party and one of th em offered the opportunity to try an illegal drug. How likely is it that you would do each of the follo wing: (1) Take it and try it, (2) Tell them no thanks, and (3) Leave the situa tion. Participants responded to each choice using the following scale: 1 = not at all likely to 7 = very likely We reverse scored responses to the second and third options before summing the three responses ( = .80; Gibbons et al., 1998). Risk behavior Participants reported the frequency with which they have used an illegal drug in their lifetime ( 0 = none, 1= 1 time, 2 = 2-5 times, 3 = 6-10 times, 4 = more than ten times ) and during the past month (1 = None of the time 2 = once or twice 3 = a few times 4 = fairly often, 5 = every/almost every day ). Due to the high correlation between these two items (i.e., r = .57, p < .01), the hypothesized model was only te sted with lifetime frequency of drug use as the outcome variable.
27 CHAPTER 3 RESULTS The study data were analyzed using the MP LUS statistical packag e. Means, standard deviations, and correlation coefficients for the model variables are presented in Tables 1 and 2. Table 3 presents the frequencie s for both the lifetime drug use a nd drug use in the past month. The data are positively screwed, with most particip ants reporting no drug use in the past month nor during their lifetime. Because both out come measures were highly correlated ( r = .57, p < .001), the model was only tested with th e lifetime drug use outcome variable. The correlations between variab les were consistent with pr evious findings. Religiosity was correlated with illicit drug use (Donahue, 1995; Wallace & Forman, 1998), as were most of the proximal and distal features of religiosity (Gibbons, Gerrard, Cleveland, Wills, & Brody, 2004). Because the predicted relationships between variables were largely supported by the correlation matrix, the data were appropriate for structural equation modeling. All models tested were observed variable path models. Two model constructs, Cognitions and Peer/Adult Models, had multiple indicators. For Peer/Adult Models the indicators were a composite of four items that measured adult models ( = .74) and a composite of the three items that measured exposure to peer models ( = .55). For Cognitions, the in dicators were the attitude item, action-specific item, four items measuri ng social norms, and the prototype items. The indicators were subjected to an explanatory factor analysis (EFA) to determine if it was appropriate to combine each set of indicators to fo rm the two model construc ts. The results of the EFAs supported combining the indicators for Cognitions and the indicators for Peer/Adult Models. For both the Cognitions EFA and the Peer /Adult Models EFA, the analyses suggested a one factor models were the best solution a nd indicators had factor loadings above .40.
28 Model 1 tested the hypothesized observed vari able path model (Figure 2). Goodness of fit indices for Model 1 fell below criteria that indicate good model fit (Table 4). Thus, the modifications indices were examined to dete rmine modifications to the model that would increase model fit. Decisions to modify the m odel were based on modification indices as well as theoretical concerns. Model 2 tested the hypothesized observed variable path model with the addition of a path from opportunities to cognitions. The modification indices suggested that the inclusion of this path would increase model fit. Mo re importantly, it made conceptual sense to include this path. The perceived ease/difficulty of obtaining drugs would likely influen ce cognitions about drug use. The direction of the path coefficient indicated that greater percei ved opportunity leads to more favorable cognitions about drug use. Perhaps people believe that if dr ugs are easy to obtain, they must not be that bad. After all, if they we re bad, then authorities w ould take more action to curb their availability. The inclusion of the path from opportunities to cognitions improved model fit (Table 4), but further modifications we re needed to meet goodness of fit guidelines. Model 3 tested the same paths represented in Model 2 with the addi tion of a path from Peer/Adult Models to Behavior al Willingness. Again, this path was added based on the modification indices and theoretical considerations Recall that Peer/Adult Models refers to the extent to which people have adults and peers that who they can look up to, aspire to be like, and who discourage engagement in risky behavior. For some people, having adults and peers that discourage risky behavior may be all that is n eeded discourage willingness to engage in risky behavior, without any cognitive elaboration. The addition of the path from Peer/Adult Models to Behavioral Willingness improved model fit to mini mum standards (Table 4), but one additional modification to the model was made to achieve good model fit.
29 The final model, Model 4, tested the same pa ths represented in Model 3 with the addition of a path from Cognitions to Risk Perceptions (F igure 3). To the extent that people hold negative cognitions regarding drug use (e.g., they have a negative attitude, perceive few social norms that support drug usage, etc.), they are likely to perceive greater risk of experiencing negative outcomes (e.g., getting addicted to drugs, suffering negative health outcomes, getting arrested for drug use) related to drug use. Good model fit was achieved by adding this path from Cognitions for Risk Perceptions (Table 4).
30 Table 3-1. Correlations for model constructs. Construct 1 2 3 4 5 6 7 8 1. Religiosity 2. Religious Coping .84** 3. Prescriptions for Behavior .10 .08 4. Peer/Adult Models .23** .13* .35** 5. Opportunities -.12* -.13* -.08 .07 6. Cognitions -.16** -.09 -.38** -.19** .44** 7. Risk Perceptions -.01 .05 .13* -.04 -.19** -.32** 8. Behavioral Willingness -.21** -.13* -.34** -.29** .36** .65** -.29** 9. Drug Use -.23** -.20** -.29** -.14* .39** .52** -.30** .55** *p .05, ** p .01
31 Table 3-2. Means and standard de viations for model constructs. Construct Mean SD Religiosity 22.98 10.97 Religious Coping 14.25 8.13 Prescriptions for Behavior 6.21 1.25 Peer/Adult Models 39.93 6.40 Opportunities 4.19 1.96 Cognitions 51.82 16.54 Risk Perceptions 83.47 70.32 Behavioral Willingness 6.33 3.92 Drug Use .96 1.53
32 Table 3-3. Frequencies for lifetime illegal drug use and drug use in the past month. Items 0 = none 1 = 1 time 2 = 2-5 times 3 = 6-10 times 4 = more than 10 times Please indicate the number of times you have used an illegal drug. 175 11 22 7 41 0 = none of the time 1 = once or twice 2 = a few times 3 = fairly often 4 = everyday or almost everyday During the past month, how often have you used illegal drugs? 243 16 6 4 5 NOTE: 18 PARTICIPANTS DID NOT ANSW ER THE LIFETIME DRUG USE ITEM.
33 Table 3-4. Summary statistics of the religiosity and risk behavior stru ctural equation models Model df X2 RMSEA (90% CI) p-value for close fit RMSEA CFI SRMR TLI 1 22 134.43* .14 (.12 .16) <.01 .85 .11 .76 2 21 75.69* .10 (.08 .12) <.01 .92 .07 .88 3 20 62.24* .09 (.06 .11) <.01 .95 .06 .90 4 19 43.65* .07 (.04 .09) .11 .97 .04 .94 N = 272. RMSEA = root mean square error of approximation; CFI = comparative fit index; SRMR = standardized root mean square residual; TLI = Tucker-Lewis Index.
34 Figure 3-1. Model 1: The hypothe sized model. Significant paths appear as solid lines, nonsignificant paths are dashed lines. .44 -.19 -.08 .34 .23 -.12 Opportunities Peer/Adult Models Cognitions: Attitudes, Norms, Beliefs, & Prototypes .84 -.12 -.11 .03 Religiosity Religious Coping Prescriptions for Behavior Risk Behavior (Drug Use) Upstream Features of Religiosity Downstream Features of Religiosity -.38 .23 -.11 .63 Risk Perceptions Behavioral Willingness
35 Figure 3-2. Model 4: The final model. Significant paths appear as solid lines, non-significant paths are dashed lines. .43 -.29 -.18 .43 -.06 -.12 .34 .23 -.12 Opportunities Peer/Adult Models Cognitions: Attitudes, Norms, Beliefs, & Prototypes .84 -.12 -.11 .03 Religiosity Religious Coping Prescriptions for Behavior Risk Behavior (Drug Use) Upstream Features of Religiosity Downstream Features of Religiosity -.31 .22 -.11 .57 Risk Perceptions Behavioral Willingness
36 CHAPTER 4 DISCUSSION This research was designed to provide an ini tia l test of a model of religiosity and risk behavior. We designed the model to predict a variety of risk beha viors including drinking alcohol, having unprotected sex, and smoking cigarettes. The current investigation examined the relationships between religiosit y, the upstream and downstream feat ures of religiosity, and illicit drug use. Consistent with our predictions, great er religiosity as measured by the Religious Commitment Inventory (RCI) corresponded to less illicit drug use. Also consistent with our predictions, most of the paths through which we hypothesized reli giosity would relate to risk behavior were significant. Howe ver, some modifications to the model were necessary to achieve good model fit. I made three modifications to the model based on goodness-of-fit i ndices and theoretical considerations. The final model included the pr oposed model paths with the addition of three new paths: 1) Opportunities to Cognitions; 2) Peer/Adult Models to Behavioral Willingness; and 3) Cognitions to Risk Perceptions. Although thes e paths were not predicted in the hypothesized model, they are justifiable bot h on theoretical and empirical gr ounds. Future tests of the model with different risk events will help us to determine whether these additional paths should become a permanent fixture of the model. Analysis of the hypothesized model revealed th at all but two of the predicted paths were significant. The first non-significan t path was the path from religi osity to prescr iptions against drug use. We can think of two reasons for the null effect. First, peopl e likely receive the messages condemning the use of illegal drug from other sources such as the media, friends, parents, and teachers, and thus there may be no variance unique to religiosity. It is noteworthy that the mean for the prescription item ( M = 6.21 on a seven point scale) suggests that most
37 participants reported that they had heard the message that using illicit drugs is wrong. Of the 270 total participants, 79% reported high agreement with the item (i.e., responded with a 6 or 7) I have been taught that using drugs is wrong, 20% reported moderate agr eement with the item (i.e., responded with 3, 4, or 5) and only 1% reported low agreement with the item (i.e. responded with a 1 or 2). These frequencies suggest that there is some variation in exposure to, or perhaps endorsement of, this message among colle ge students, but it is mostly due to outliers in the data. Second, the items may have been poorly worded and thus did not captu re the construct of behavioral prescriptions. We created and tested the prescription item for the first time in this study. Perhaps participants dont regard the transm ittal of information in a religious context or from adults in their religious community as t eaching. Rather, they may have a limited view of teaching as something that occurs only at schoo l. In future studies we will refine the measurement of prescriptions for behavior constr uct by including items that assess the different sources of exposure to prescriptions. It is quite possible that young people are exposed to competing prescriptions regardi ng risk behavior. For example, music videos may endorse drug use, while parents and religi ous leaders discourage it. The second non-significant path in the hypothe sized model was the path from Peer/Adult Models to Cognitions. Our data suggest that e xposure to peers and adults who can serve as models for non-risky behavior and discourage ri sky behavior was not related to cognitions regarding drug use. However, this hypothesized pa th becomes significant in the final model. The switch from a nonsignificant path to a significant path following the addition of three new paths to the final model (i.e., the paths from peer/adult models to behavioral wi llingness, cognitions to
38 risk perceptions, and opportunitie s to cognitions) suggests a s uppressor effect. Say in one sentence what this means (i.e., what is being suppressed). Several additions findings deserve men tion. First, we observed a high correlation between religiosity and our m easure of religious coping ( r = .84) suggesting that the items represented by the two measures were tapping the same underlying construct. We are currently looking for a new measure of coping to addr ess this problem. Second, we observed a high correlation between cognitions and behavioral willingness ( r = .65). This result is not surprising. Previous researchers have found strong relationships between the cognitions constructs and behavioral willingness (Gerrard, Gibbons, Re is-Bergan, Trudeau, Vande Lune, & Buunk, 2002; Gibbons et al., 1998). In additi on, the measurement of these constructs was adapted from measures used in previous studies. Thus, the items measuring the cogni tions constructs and behavioral willingness have undergone multip le modifications and are more precise measurements of the constructs. There are additional limitations to the curren t investigation. First, participants in the current study were college students, primarily between the ages of 18 and 21. These late adolescents/emerging adults likely differ fr om younger adolescents (e.g., they have more opportunity to engage in risk be havior due to less parental m onitoring). Second, the current study only examined drug and not other risky behaviors. It is unknown how the results will look when other risk events (e.g., alcohol us e, unprotected sex) are examined. For example, it may be that opportunity is more important in predicting drug use than in pred icting cigarette smoking because illicit drugs are more di fficult to obtain. Third, all da ta are self-reports. Perhaps participants are not accurately reporting their behavior and perceptions. Participants may not accurately report their drug use be cause of fear of the consequences even though we assured
39 participants repeated that there responses woul d remain confidential. Also, participants may misreport their behavior and perceptions because they simply do not remember that information accurately. Fourth, the cross-sectional natu re of the data limits our abil ity to draw conclusions about the ordering of events in the model. In addition, the nature of the dependent measure limits the ability to conclude that religiosity and the upstream and downstream features of religiosity precede drug use in time. The dependent measur e for the current study was lifetime drug use. The drug use participants reported could have occurred recently or when they were younger (e.g., in high school). However, participants re ported their current re ligiosity and current perceptions of the upstream and do wnstream features of religiosity (e.g., attitudes, opportunities). Thus, the behavior reported in the outcome meas ure could have occurred before other model components. It is possible that drug use influenced their perceptions of religion and other model constructs rather than reli giosity and its features pr edicted their drug use. Our findings, and the limitations we addresse d, suggest important directions for future research. First, we are interested in testing the model with younger adolescents. As stated previously, younger adolescents may differ in si gnificant ways from young college students and these differences may produces differences in th e predictive capacity of the model. Second, we plan to multiple measurements of model constr ucts so that we can examine the effects of religiosity and its upstream and downstream featur es on risk behavior ov er time. Next, we are interested in examining whether our model predicts for other risk events. Finally, we plan to increase the number of items that measure pres criptions to assess the source of prescriptions against risk behavior (e .g., media, parents).
40 For years researchers ha ve found that religious adolescents are less likely to engage in risk behavior (e.g., Amey, Albrecht, & Mill er, 1996; Dunn, 2005; Nonnemaker et al., 2003; Steinman & Zimmerman, 2004; Wallace & Forman, 1998), but few studies have examined why. The theoretical model we present is an important first step in learning w hy religious adolescents are disinclined to engage in risky behavior Our data supported the hypothesized theoretical model with minor alterations. However, some th eorized paths in the model are more predictive (i.e., have larger indirect effects) than are othe r paths. The Religiosity and Risk Behavior Model does not specify a direct path be tween religiosity and drug use. Ra ther, it specifies links between intervening upstream and downstrea m features of religiosity that account for the relationship between religion and drug use. These indirect effects are calculated by multiplying the coefficients along each pathway linking religiosit y to drug use. For exam ple, the size of the indirect effect associated with the link of religiosi ty to religious coping to drug use was .25. While this was the largest indirect path, the problematic measurement of religious coping limits our ability to draw conclusions from this effect. Two model paths stand out as most predictiv e of drug use (excluding those paths that include variables with measurement problems). First, the path from religiosity to opportunities to drug use suggests that the lower illicit drug use among religious adolescents is partially explained by their having less opportunity to us e drugs. The implication is that drug use can be among adolescents by restricting the access or to us e drugs. Second, the path from religiosity to peer/adult models to cognitions to behavioral willingness to drug use sugg ests that religious adolescents report having more peer and adult models that they can look up to and that discourage the use of drug use. In addition, exposu re to these models influences cognitions about drug use (e.g., leads to more nega tive attitudes, more negative evaluation of the type of person
41 who uses drugs) that then leads to a decreased willingness to use drugs and less actual drug use. The implication is that providing youth with peer/adult models who discourage drug use can foster negative toward illicit drugs. Consistent with this notion is evidence that adolescents who regularly interact with a positive model are 46% le ss likely to begin using drugs, compared with adolescents who are on a wait list for the ment oring program (Tierney, Grossman, Resch, 1995). In summary, the current study is the first test of the Religiosi ty and Risk Behavior model. Although the study had some limitations, overall the data support the hypothesized model. Religious adolescents were less likely to re port using illicit drugs and their lower drug use appears to result primarily from greater exposure to peer/adult models who discourage drug use and less opportunity to use illicit drugs.
42 APPENDIX A CONSENT FORM I will answer several questionnaires in this s tudy. These questionnaires will ask about my feelings towards different behaviors. I will r eceive 2 experimental credits for completing the questionnaires today. Depending on the speed of your computer conn ection and the speed at which you answer questions, your participation to day may take up to one hour. Time Required : 1 hour Risks and Benefits : I will benefit by learning about research. There are no risks. Compensation: I will receive 2 credits for participation. Confidentiality : My responses will be confidential to th e extent provided by the law. I will be assigned a code number, and my responses will be stored in a computer according to the code number and not by my name. As such, my name will not be associated with my responses and will not be used in any report. Moreover, all da ta will be analyzed by group averages and not by individual responses. Voluntary Participation & Right to Withdraw : I understand that my participation in this study is voluntary. There is no penalty for not partic ipating. I have the right to withdraw from the study at any time without consequence. Whom to Contact if You have Questions about the Study : James A. Shepperd, Faculty Advisor, Dept. of Psychology, University of Florida, 392-0601 x 248. Wendi Malone, Principal Invest igator, Dept. of Psychology, Univ ersity of Florida, 392-0601 x 261. Whom to Contact about Your Rights as a Research Participant in the Study : UFIRB Office, Box 112250, University of Fl orida, Gainesville, FL 32611-2250; ph. 392-0433. By signing below I acknowledge that I have read the above information and agree to participate in this study. ______________________________ ____________ Signature of Research Participant Date
43 APPENDIX B DEMOGRAPHIC QUESTIONNAIRE Please answ er the following demographic questions. 1. Age:_____ 2. Sex: _____Female _____Male 3. Ethnicity: a. African American b. Arab American c. Asian American d. Caucasian e. Hispanic f. Native American g. Other:_________________ (please specify) 4. Class rank: a. Freshman b. Sophomore c. Junior d. Senior e. Grad Student f. Other:__________ (please specify) 5. What is your religious affiliation? a. Agnostic b. Atheist c. Christian d. Islamic e. Jewish f. Other. Please specify _______________________________ 6. How often do you attend religious services (on average over the past year)? a. Once a week or more b. Two or three times a month c. Once a month d. Only on important holidays e. Never 7. On average, how many hours per month do you spend participating in activities (for example, Bible studies) associated with your religious institution? a. More than one hour per week b. One hour per week c. One to three hours per month d. Less than one hour per month
44 e. None 8. On average, how often did you attend re ligious service thr oughout your childhood? a. Once a week or more b. Two or three times a month c. Once a month d. Only on important holidays e. Never
45 APPENDIX C RELIGIOUS COMMITMENT INVENTORY (RCI-10) Instructions: Read each of the following statements. Using the drop down boxes, indicate the extent to which each statement is true of you. Not at all Somewhat Moderately Mostly Totally true of me true of me true of me true of me true of me 1 2 3 4 5 1. I often read books and magazines about my faith. 2. I make financial contributions to my re ligious organization. 3. I spend time trying to grow in understa nding of my faith. 4. Religion is especially important to me becau se it answers many questions about the meaning of life. 5. My religious beliefs lie behind my whole approach to life. 6. I enjoy spending time with others of my religious affiliation. 7. Religious beliefs influence all my dealings in life. 8. It is important to me to spend periods of time in private religious thought and reflection. 9. I enjoy working in the activities of my religious affiliation. 10. I keep well informed about my local religious group and have some influence in its decisions.
46 APPENDIX D PEER/ADULT MODELS Please rate the following items on a scale from 1 = strongly disagree to 7 = strongly agree. 1. I know adults who care about me. 2. I have peers who care about me. 3. The adults I know discourage or disapprove of risky behavior. 4. I feel I have many adults that I look up to. 5. The peers I spend time with discourage or disapprove of risky behavior. 6. I know many adults that I admi re and aspire to be like. 7. I know many peers that I admire and aspire to be like.
47 APPENDIX E RELIGIOUS COPING STRATEGIES Please think about a stressful event in your past. P lease read the statements listed below and for each statement indicate to what extent each of the following was involved in your coping with the event. Please answer the following items on a scale from 1 = not at all to 7 = a great deal. 1. Found the lesson from God in the event. 2. Sought support from clergy. 3. Attended religious services or pa rticipated in religious rituals. 4. Participated in religious groups (e.g., support groups, pray er groups, Bible studies). 5. Sought support from other members of the religious institution.
48 APPENDIX F SUBJECTIVE NORMS Rate the follow ing items on a scale from 1 = none to 7 = almost all 1. How many people your age do drugs or have used drugs? 2. How many of your friends do drugs or have used drugs? Rate the following items on a scale from 1 = have a strong negative reaction and tell you to stop to 7 = encourage you to continue 3. How do you think your friends would respond if they thought you did drugs? 4. How do you think your parents would respond if they thought you did drugs?
49 APPENDIX G PROTOTYPES Prototypes The questions below concern your im ages of peopl e. What we are interested in here are your ideas about typical members of different groups. For example, we all have ideas about what typical movie starts are like or what the typical grandmother is like. When asked, we could describe one of these imageswe might say we thi nk the typical movie star is pretty and rich, or that the typical grandmother is sweet and frail. We are not saying that all movi e stars or all grandmothers are exactly alike, but rather that many of them share certain characteristics. Please indicate the extent to which each of the following adjectives descri bes the type of person (your age) who uses drugs. Please rate adjectives on a scale from 1 = not at all to 7 = very. smart, confused, popular, immature "cool" (sophisticated), self-confident, i ndependent, careless, unattractive, dull (boring), considerate and self-centered
50 APPENDIX H RISK PERCEPTIONS In this section we ask you to im agine YOURSELF in situations. You personally might never be in these situations. But try to estimate the chance th at these things would happen to YOU if you were in these situations, by usi ng any number from 0% to 100%. Imagine that you are at a party and someone offe rs you an illegal drug. You decide to try it. What is the chance that you will overdose on the drug? What is the chance that you will get arrested for using the drug? What is the chance that you w ill get addicted to the drug?
51 REFERENCES Adlaf, E. M., & Sm art, R. G. (1985). Drug use and religious affiliation, feelings, and behaviour. British Journal of Addiction, 80, 163-171. Ajzen, I. & Fishbein, M. (1980). The predicti on of behavior from attitudinal and normative variables. Journal of Experime ntal Social Psychology, 6, 466-487. Albrecht, S. L., Chadwick, B.A., & Alcorn, D. S. (1977). Religiosity and deviance: Application of an Attitude-Behavior Conti ngency Consistency Model. Journal for the Scientific Study of Religion, 16, 263-274. Amey, C. H., Albrecht, S. L., & Miller, M. K. (1 996). Racial differences in adolescent drug use: The impact of religion. Substance Use and Misuse, 31, 1311-1332. Baumrind, D. (1987) A developmental perspectiv e on adolescent risk taking in contemporary America. New Directions fo r Child Development, 37, 93-125. Blanton, H., Gibbons, F. X., Gerrard, M., Conger, K. J., & Smith, G. E. (1997). Role of family and peers in the development of prototypes a ssociated with substa nce use. Journal of Family Psychology, 11, 271-288. Burkett, S., & White, M. (1974). Hellfire and delinquency: Another look. Journal for the Scientific Study of Religion, 13, 455-462. Burkett, S. R., & Warren, B. O. (1987). Religiosity, peer associ ation, and adolescent marijuana use: A panel study of underlying cau sal structures. Criminology, 25, 109-131. Centers for Disease Control and Prevention. (2004). Youth risk behavior surveillance United States, 2003. Morbidity and Mortality Weekly Report, 53(SS02), 1-96. Cochran, J. K., & Akers, R. L. (1989). Beyond hellfir e: An exploration of th e variable effects of religiosity on adolescent marijuana and alcohol use. Journal of Research in Crime and Delinquency, 26, 198-225. Cornwall, P. H., Albrecht, S. L., Cunningham, P. H., & Pitcher, B. L. (1986). The dimensions of religiosity: A conceptual model with an empi rical test. Review of Religious Research, 27, 226-244. Donahue, M. J. (1995). Religion an d the well-being of adolescents. Journal of Social Issues, 51, 145-160. Dunn, M. S. (2005). The relationship between reli giosity, employment, and political beliefs on substance use among high school seniors. Journal of Alcohol and Drug Education,49, 7388. DuRant, R. H., Pendergrast, R., & Seymore, C. (1990). Sexual behavior among Hispanic female adolescents in the United States. Pediatrics, 85, 1051-1058.
52 Fielding, J. E., & Williams, C. A. (1991). Adolescent pregnancy in the United States: A review and recommendation for clinicians and research needs. American Journal of Preventive Medicine, 7, 47-52. Gerrard, M. Gibbons, F. X., Reis-Bergan, M., Trudeau, L., Vande Lune, L. S., & Buunk, B. (2002). Inhibitory effects of drinker and nondrinker prototypes of adolescent alcohol consumption. Healthy Psychology, 21, 601-609. Gibbons, F. X., Gerrard, M., Blanton, H., & Russe ll, D. W. (1998). Reasoned action and social reaction: Willingness and intention as independent predictors of health risk. Journal of Personality and Social Psychology, 74, 1164-1180. Gibbons, F. X., Gerrard, M., Boney-McCoy, S. ( 1995). Prototype percepti on predicts (lack of) pregnancy prevention. Personality and Social Psychology Bulletin, 21, 85-93. Gibbons, F. X. Gerrard, M., Cleveland, M. J., Wills, T. A., & Brody, G. (2004). Perceived Discrimination and Substance Use in African American Parents and Their Children: A Panel Study. Journal of Personality and Social Psychology, 86, 517-529. Gibbons, F. X., Gerrard, M., Ouelette, J., & Bu rzette, B. (1998). Cognitive antecedents to adolescent health risk: Discriminating betw een behavioral intention and behavioral willingness. Psychology and Health, 13, 319-340. Gibbons, F. X., Helweg-Larsen, M., & Gerrard, M. (1995). Prevalence estimates and adolescent risk behavior: Cross-cultural differences in social influence. Journal of Applied Psychology, 80, 107-121. Halpern-Felsher, B. L., Millstein, S. G., Ellen, J. M., Adler, N. E., Tschann, J. M., Biehl, M. (2001). The role of behavioral experience in judging risks. H ealth Psychology, 20, 120126. Higgins, P.C., & Albrecht, G. L. (1977). Hellfire and delinquency revisited. Social Forces, 55, 952-958. Hill, P. C., & Hood, R. W., Jr. (Eds.). (1999) Measures of Religiosity. Birmingham, AL: Religious Education Press. Hirschi, T., & Stark, R. (1969). Hellfire and delinquency. Social Problems, 17, 202-213. Jensen, G., & Erickson, M. (1979). The religious factor and de linquency: Another look at the Hellfire hypothesis. In R. Withnow (Ed.) The Religious Dimension (pp. 157-177). New York: Academic Press. Johnston, L. D., OMalley, P. M., & Bachman, J. G. (2002). Monitoring the Future national results on adolescent drug use: Overview of key findings, 2001. (NIH Publication No. 025105). Bethesda, MD: National Institute on Drug Abuse.
53 Lorch, B. R., & Hughes, R. H. (1985). Religion a nd youth substance use. J ournal of Religion and Health, 24, 197-208. Miller, L., Davies, M., & Greenwald, S. (2000). Religiosity and substance use and abuse among adolescents in the National Comorbidity Survey. Journal of the American Academy of Child and Adolescent Psychiatry, 39, 1190-1197. Moore, S., & Parsons, J. (2000). A research age nda for adolescent risk-taking: Where do we go from here? Journal of Adolescence, 2, 371-376. Mullen, K., & Francis, L. J. (1995). Religios ity and attitudes toward drug use among Dutch school children. Journa l of Alcohol and Drug Education, 41, 16-25. Nonnemaker, J. M., McNeely, C. A., & Blum, R. W. (2003). Public and private domains of religiosity and adolescent health risk beha viors: Evidence from the National Longitudinal Study of Adolescent Health. Social Science & Medicine, 57, 2049-2054. Pargament, K. I. (1997). The psychology of religion and coping. New York: Guilford. Powell, K. B. (1997). Correlates of violent and nonviolent behavior among vulnerable inner-city youths. Family & Community Health, 20, 38-47. Powers, G. E. (1967). Prevention through religion. In W. E. Amos & C. F. Wellford (Eds.), Delinquency prevention: Theory and practi ce (pp. 99-127). New York: Prentice Hall. Resnick, M. D., Bearman, P. S., Blum, R. W., Baum an, K. E., Harris, K. M., Jones, J., et al. (1997). Protecting adolescent from harm: Findings from the National Longitudinal Study on Adolescent Health. Journal of the American Me dical Association, 278, 823-832. Rew, L., & Wong, Y. J. (2006). A system atic review of associations among religiosity/spirituality and adol escent health attitudes and beha viors. Journal of Adolescent Health, 38, 433-442. Shedler, J. & Block, J. (1990). Adolescent dr ug use and psychological health: A longitudinal inquiry. American Ps ychologist, 45, 612-630 Stark, R. (1996). Religion as context: Hellfire and delinquency one more time. Sociology of Religion, 57, 163-173. Steinman, K. J., & Zimmerman, M. A. (2004). Religious activity and risk behavior among African American adolescents: C oncurrent and developmental e ffects. American Journal of Community Psychology, 33, 151-161. Takyi, B. K. (2003). Religion and women's health in Ghana: Insights into HIV/AIDs preventive and protective behavior. Social Science and Medicine, 56, 1221-1234. Tierney, J. P., Grossman, J. B., & Resch, N. ( 1995). Making a difference: An impact study of big brothers/bit sisters. Philadelphi a: Public/Private Ventures.
54 Tittle, C. R., & Welch, M. R. (1983). Religiosity and deviance: Toward a contingency theory of constraining effects. Social Forces, 61, 653-682. Tuinstra, J., Groothoff, J. W., Van Den Heuvel, W. J. A., & Post, D. (1998). Socio-economic differences in health risk be havior in adolescence: Do th ey exist. Social Science & Medicine, 47, 67-74. Wallace, J. M., Jr., & Forman, T. A. (1998). Reli gions role in promoting health and reducing risk among American youth. Health Education and Behavior, 25, 721-741. Weden, M. M., & Zabin, L. S. (2005). Gender an d ethnic differences in the co-occurrence of adolescent risk behaviors. Ethnicity & Health, 10, 213-234. Worthington, E. L., Jr., Wade, N. G., Hight, T. L., Ripley, J. S., McCullough, M. E., Berry, J. W., et al. (2003). The Religious Commitment I nventory-10: Development, refinement, and validation of a brief scale for research a nd counseling. Journal of Counseling Psychology, 50, 84-96.
55 BIOGRAPHICAL SKETCH Wendi A. Malone is a third year graduate student in social psychology. In 2006, she earned a Master of Science degree in psychology from Augusta State Universi ty (Augusta, Georgia). She received her bachelors degree in ps ycholog y in 2004 from William Penn University (Oskaloosa, Iowa).