JOURNAL OF AMERICAN COLLEGE HEALTH, VOL. 58, NO.

Associations Between the Five-Factor Model of Personality and Health Behaviors Among College Students
Douglas A. Raynor, PhD; Heidi Levine, PhD

Abstract. Objective: In fall 2006, the authors examined associations between the five-factor model of personality and several key health behaviors. Methods: College students (N = 583) completed the American College Health AssociationNational College Health Assessment and the International Personality Item Pool Big Five short-form questionnaire. Results: Highly conscientious individuals were more likely to wear seat belts, utilize alcohol-related harm reduction, exercise, get enough sleep, and consume fruits and vegetables. They were also less likely to smoke cigarettes, consume alcohol, and binge drink. Highly extraverted individuals were more likely to smoke cigarettes, consume alcohol, binge drink, and have multiple sexual partners, and they were less likely to engage in alcohol-related harm reduction, use condoms, and get enough sleep. Conclusions: These findings are supportive of a growing body of evidence indicating that conscientiousness and extraversion are robust concomitants of health behaviors among college students. Keywords: college students, five-factor model of personality, health behavior, personality traits

any contemporary public health problems are highly preventable. Regularly engaging in preventive health behaviors (eg, physical activity) and avoiding deleterious health behaviors (eg, cigarette smoking) reduce the risk of acute injury or illness, several chronic diseases, and premature mortality.1 Given that a significant number of college students are nonadherent to multiple health behavior goals delineated in Healthy Campus 2010,2,3 there exists a strong need to identify predictors of their preventive and risky health behaviors.
Dr Raynor is with the Department of Psychology at the State University of New York at Geneseo. At the time of the study, Dr Levine was with the Lauderdale Center for Student Health and Counseling at the State University of New York at Geneseo. She is now with the Office of Student Affairs at Cornell College, Vernon, Iowa. Copyright 2009 Heldref Publications 73

The five-factor model (FFM) of personality has emerged as a promising predictor of health behaviors. The FFM is a parsimonious taxonomy that classifies all personality dimensions into 5 broad domains, commonly labeled openness, conscientiousness, extraversion, agreeableness, and neuroticism.4 Openness refers to being intelligent, imaginative, curious, flexible, and broad-minded. Conscientiousness refers to striving for competence and achievement, and being selfdisciplined, orderly, reliable, and deliberative. Extraversion refers to enjoying the company of others, and being active, talkative, assertive, and seeking stimulation. Agreeableness refers to being courteous, good-natured, cooperative, tolerant, and compassionate rather than antagonistic. Neuroticism refers to easily experiencing unpleasant and negative emotions, such as fear, anxiousness, pessimism, sadness, and insecurity. Extensive evidence supports the reliability and construct validity of these Big Five dimensions of personality across ages and cultures.5,6 Most studies examining the predictive utility of the FFM for undergraduates have targeted isolated health behaviors, often with mixed findings. Of the Big Five factors, conscientiousness and extraversion have received the most consistent support, with high conscientiousness associated with increased exercise;7 improved sleep sufficiency, sleep quality, and earlier sleep schedules (ie, wake-up and retiring times);8 safer sexual behavior;9 decreased risky driving behavior;10,11 decreased alcohol consumption and disorders;1214 and decreased cigarette smoking.14 Different aspects of extraversion have been associated with both preventive and risky health behaviors. Extraversion, particularly the facet of activity, is positively associated with exercise engagement,7 but high levels of extraversion and other facets (eg, sensation seeking) have been associated with increased alcohol consumption, alcohol-related problems, binge drinking,1517 cigarette smoking,18 risky

Raynor & Levine

driving behavior,19,20 and risky sexual behavior.9 However, several studies failed to replicate significant associations among conscientiousness, extraversion, and the aforementioned health behaviors.7,12,13,1922 Only modest support exists for the utility of neuroticism in predicting various health behaviors. High neuroticism has been associated with decreased exercise adherence;7 increased alcohol consumption, alcohol-related problems, and binge drinking;12,13,17,23 increased cigarette smoking;18,22 and decreased sleep quality and sleep efficiency.8 However, several studies have failed to replicate the aforementioned associations involving neuroticism,7,14,15 as neuroticism was unrelated to risky driving10,19 and sexual behaviors.9 Agreeableness was also unrelated to level of exercise behavior,7 sleep quality and efficiency,8 and alcohol-use and disorders.13 However, high overall agreeableness did predict decreased alcohol use disorders,12 and 2 subcomponents of agreeablenessstraightforwardness and tendermindednesswere inversely related to alcohol use and alcohol disorders.13 Finally, there is a dearth of studies reporting significant associations between openness and health behaviors.7,8,10,13,19 A notable exception exists with 1 study reporting a positive association between openness and exercise engagement.24 A significant limitation in the FFM health behavior literature is that almost all studies have examined health behaviors independently. Given the ample evidence showing that health behaviors tend to coexist within individuals,25 it is imperative to examine the extent to which personality traits influence a range of seemingly disparate health behaviors. Only a few international studies have examined associations between the complete FFM and multiple health behaviors of undergraduates. In a large sample of 1,184 collegeaged men and women in Spain, conscientiousness was a particularly strong predictor of health behaviors, including decreased smoking and alcohol consumption, and increased exercise and healthy eating habits.26 Less potent but still significant, sex-specific correlations were also observed: (1) Increased neuroticism, extraversion, and openness were linked with unhealthy habits in women; (2) increased agreeableness was linked with healthy habits in women; and (3) increased neuroticism was linked with unhealthy habits in men. In a small sample of 150 college students in the United Kingdom, agreeableness and conscientiousness were positively associated with a 29-item preventive health behavior scale. Results also indicated interactive effects such that high agreeableness/high conscientiousness and low openness/high conscientiousness predicted greater preventive behavior.27 Finally, in a sample of 683 Swiss university students, regardless of level of extraversion and neuroticism, high conscientiousness was associated with reductions in smoking cigarettes, drinking alcohol, drunkenness, drunk driving, cannabis use, and number of sexual partners. Conversely, students with low conscientious and either high extraversion or high neuroticism were more likely to engage in several of the aforementioned risky health behaviors.28
74

Another important limitation in the personality-health behavior literature is that associations among the FFM and several critical health behaviors delineated in Healthy Campus 2010 guidelines,3 including sleep, diet, sexual risk taking, and alcohol-related harm reduction, have received minimal empirical attention. A careful review revealed only 1 published study focusing on sleep behavior,8 another addressing sexual risk taking,9 no studies focusing on diet and alcohol-related harm reduction, and many including only 1 or 2 of the Big Five as predictors. Another limitation is that a wide range of health behaviors or risks were assessed with surveys lacking psychometric validation, which may have contributed to the heterogeneity of results. Finally, there is growing evidence that male students, Greek-affiliated students, and intercollegiate athletes are more likely to engage in certain unhealthy behaviors.2,29,30 Unfortunately, the majority of personality-health behavior studies have failed to determine whether personality traits explain variance beyond that accounted for by these important demographic and group membership variables. We attempted to overcome several of the aforementioned limitations by examining putative associations between the complete FFM of personality and several key preventive and deleterious health behaviors delineated in Healthy Campus 2010.3 The American College Health Association National College Health Assessment (ACHA-NCHA) and the short-form International Personality Item Pool (IPIP) Big Five questionnaire were administered via the Internet to a sample of college students at a liberal arts college. On the basis of prior evidence, we hypothesized that conscientiousness would be associated with increased preventive health behaviors and decreased risky health behaviors. We expected extraversion to have divergent associations, predicting increased physical activity, as well as an increase in several risky behaviors. Given the modest empirical support in the literature, we did not hypothesize a priori significant links for the traits of neuroticism, agreeableness, and openness. Finally, we wanted to examine whether the predictive utility of personality traits would exist after controlling for a few preselected group membership variables, including sex, intercollegiate athletic status, and Greek affiliation. METHODS Participants A total of 2,000 undergraduate students enrolled at a small (ie, enrollment < 5,500) public liberal arts college in the northeastern United States were invited to participate in this study during the fall 2006 semester. To recruit a sufficient number of students involved in intercollegiate athletics and Greek organizations, all individuals from these 2 groups were invited to participate. The athlete and Greek-affiliated groups comprised 764 students (378 intercollegiate athletes, 471 Greek members, 85 belonging to both groups), and the remaining 1,236 invited students were selected from the general student body in a stratified random sample with equal distribution across undergraduate class years. As an
JOURNAL OF AMERICAN COLLEGE HEALTH

Personality and Health Behaviors

incentive to participate, 10 randomly selected students were awarded a $25 bookstore gift certificate at the conclusion of the study. Participation in this study was voluntary, all students provided informed consent, and the colleges Institutional Review Board approved this study. Data were collected from 603 participants, representing a 30% response rate. Twenty participants were dropped from this sample because of missing data (> 15% of survey items), resulting in a final sample of 583 participants. The sample comprised 117 intercollegiate athletes (20.1%) and 95 Greek-affiliated members (16.3%). Participants identified as predominantly female (74%) and white (89%), with 0.3% identified as black, 1.7% as Hispanic/Latino, 3.6% as Asian/Pacific Islander, and 4.3% as other or having more than 1 ethnicity. In comparison to the student body, the sample was disproportionately composed of female students (overall female enrollment: 59%), intercollegiate athletes (overall enrollment: 7.6%), and Greek members (overall enrollment: 9.4%), but it was consistent with the 11% of students of color enrolled overall during the same time period. The sample was evenly distributed across class years, with approximately one-quarter in each academic class. Measures Health Behaviors The ACHANCHA Web survey2 consists of 240 questions assessing a range of health-related behaviors, problems, and beliefs and is typically completed in 2030 minutes. Six topic areas, in addition to demographic characteristics, are assessed: (1) health, health education, and safety; (2) alcohol, tobacco, and drugs; (3) sex behavior, perceptions, and contraception; (4) weight, nutrition, and exercise; (5) mental and physical health; and (6) impediments to academic performance. The structure of the questions varies, including Likert-type ratings and yes/no questions, as does the period of time about which student behaviors are assessed (eg, past year, past month, past week). Studies have demonstrated that this survey is a reliable and valid measure of college student health behavior.31 A subset of items was selected for the current study. These items measured 11 behaviors related to injury prevention (eg, Within the last school year, how often did you wear a seatbelt when you rode in a car?), exercise (eg, On how many of the past 7 days did you participate in vigorous exercise for at least 20 minutes or moderate exercise for at least 30 minutes?), sleep quality (eg, On how many of the past 7 days did you get enough sleep so that you felt rested when you woke up in the morning?), sexual behavior (eg, Within the last school year, with how many partners, if any, have you had sex [oral, vaginal, or anal]?), fruit/vegetable consumption (eg, How many servings of fruits and vegetables do you usually have per day?), and tobacco and alcohol consumption (eg, Within the last 30 days, on how many days did you use alcohol?).
VOL 58, JULY/AUGUST 2009

Personality The IPIP is a public domain collection of over 2,000 items measuring a wide range of personality characteristics. Items involve a compact verbal phrase (eg, Make plans and stick to them, Make friends easily, Feel comfortable with myself) from which individuals rate the extent to which they perceive the behavior to be like them; responses are derived from a 5-point Likert-type scale ranging from 1 (very inaccurate) to 5 (very accurate). The 50-item IPIP Big Five factor inventory was recently found to be psychometrically valid and to correlate strongly with 2 leading personality inventories (ie, the NEO-FFI Five Factor Inventory and Eysenck Personality QuestionnaireRevised Short-Form).32 Moreover, 2 other recent studies of adolescents33 and college students,34 respectively, provided evidence of nearly identical factor structure of both selfreports and observer ratings of the 50-item IPIP Big Five inventory. These studies also evaluated convergent and discriminant validity of the IPIP Big Five factors by calculating multitraitmultimethod correlations35 between factor scores from self-report and parent rating33 or peer rating34 data. The convergent validity (or monotrait-heteromethod) coefficients of the Big Five factors substantially exceeded the heterotrait-heteromethod correlations in all cases. The mean convergent validity coefficients of the factors across the 2 assessment approaches were .48 (SD = .08) for Milas et al33 and .50 (SD = .11) for Mlacic and Goldberg,34 respectively. In contrast, the mean of the heterotrait-heteromethod correlations was .01 for both studies (maximum absolute correlation = .10), and none of these correlations approximated the size of the convergent validity coefficients. Thus, there exists strong evidence for the internal reliability, factor structure, and convergent and discriminant validity of the IPIP Big Five factor inventory for both self-report and observer ratings. A factor analysis performed by Buchanan et al36 using the 50-item inventory identified 9 redundant items, the exclusion of which resulted in a more psychometrically valid and parsimonious measure. Individual scales in the 41-item IPIP Big Five inventory include openness, conscientiousness, extraversion, agreeableness, and neuroticism. Internal consistency reliabilities (coefficient ) for the 41item revised scale range from .74 (openness) to .88 (extraversion), and criterion-related validity was established with significant correlations between each of the personality factors and relevant behavioral outcome variables.36 Another study37 provided additional evidence that the 41-item version of the IPIP Big Five inventory has satisfactory internal reliability and convergent validity. Procedures Students were invited to participate in the survey via an e-mail message in mid-October 2006 that explained the project and contained an embedded URL link to the survey. Students were informed that participation was voluntary, and that their consent was indicated through completion of the survey. Students who did not respond initially received
75

Raynor & Levine

2 reminders at approximately 10-day intervals. A unique ID was assigned to each student invited to participate in the study to prevent redundant submissions and to follow up with nonresponders. Survey responses were de-identified to ensure anonymity. Analysis All descriptive and inferential analyses were conducted using SPSS for Windows Version 14.0 (SPSS Inc, Chicago, IL). Multiple linear regression analyses were performed to determine whether the FFM personality factors were related to each of the health behaviors, with sex, intercollegiate athletic status, and Greek affiliation entered as controls. The p value for all tests was set at a conservative .01 level to minimize Type I error rate. RESULTS Description of Health Behaviors As observed in the national sample of college students completing the NCHA in the fall of 2006,2 many participants in the current sample (N = 583) adhered poorly to preventive health behaviors, and concurrently, many engaged in risky behaviors. For instance, 63% reported usually consuming 2 or fewer servings of fruits and vegetables per day, and 55% reported engaging in sufficient physical activity on 2 or fewer days per week. About 15% of the sample reported not always wearing seat belts when riding in a car during the past year. Roughly 18% reported consuming alcohol on 1030 days within the past 30 days, and 57% reported consuming 5 or more alcoholic drinks in 1 sitting during the past 2 weeks. Of the 306 participants (53% of

sample) engaging in vaginal intercourse within the past 30 days, 38% reported that they or their partner(s) rarely or never used a condom, whereas 40% reported always using a condom. Prediction of Health Behaviors The linear regression models predicting health behaviors using the FFM personality factors are presented in Tables 1 and 2. The tables organize results according to whether the health behaviors are preventive or risky in nature. In each model, sex, intercollegiate athletic status, and Greek affiliation were entered as control variables in the first block of predictors (Step 1), and the personality variables were entered in the second block (Step 2). The overall fit was statistically significant for all of the models, ps < .001, and adding the block of personality factors significantly improved model fit in each case, ps < .001. In the first set of 7 regression models, preventive health behaviors were entered as dependent variablesseat belt use, moderate/vigorous exercise, strengthening exercise, getting enough sleep, alcohol-related harm reduction, condom use, and fruits/vegetables servings (see Table 1). In the first model, conscientiousness was positively associated with seat belt use, p < .001. In the second and third models predicting physical activity, conscientiousness was associated with increased moderate/vigorous exercise, p < .001, and strengthening exercise, p = .002. Extraversion was also positively associated with strengthening exercise, p = .004. In the fourth model, conscientiousness was predictive of increased likelihood of getting enough sleep to feel rested, p = .001, whereas extraversion, p < .001, and neuroticism,

TABLE 1. Hierarchical Linear Regressions (Standardized ) of the Five-Factor Model of Personality Variables on Preventive Health Behaviors Seat belt Moderate/ Strengthening Restful Alcohol-related Condom Fruit or vegetable use (last vigorous exercise exercise sleep (past harm reductiond use (last servings (usual school year) (past 7 days) (past 7 days) 7 days) (last school year) 30 days) number per day)

Variable Step 1 Sexa Greek statusb Athletic statusc R2 Step 2 Openness Conscientiousness Extraversion Agreeableness Neuroticism R2 Adj R2 F

.13* .04 .10 .024* .01 .22* .05 .05 .03 .061* .072 6.63*

.08 .02 .53* .290* .01 .14* .06 .04 .06 .033* .314 34.26*

.10 .02 .33* .116* .00 .13* .12 .07 .04 .041* .145 13.35*

.03 .05 .10 .016 .08 .15* .18* .05 .30 .115* .119 10.81*

.13* .20* .07 .050* .04 .19* .18* .13* .02 .090* .128 11.66*

.01 .10 .06 .016 .03 .01 .25* .09 .04 .063* .066 6.13*

.10 .02 .11* .025* .14* .16* .03 .02 .04 .043* .055 5.27*

Note. For each F, dfs = 8, 574. a For sex, female (n = 428) was coded as 1; male (n = 149) was coded as 2. b Greek affiliation was coded as 0 for no (n = 488) and 1 for yes (n = 95). c Athletic status was coded as 0 for no (n = 466) and 1 for yes (n = 117). d Alcohol-related harm reduction scores ranged from 10 (low harm reduction) to 60 (high harm reduction). * p < .01, 2-tailed.

76

JOURNAL OF AMERICAN COLLEGE HEALTH

Personality and Health Behaviors

p < .001, were linked with decreased likelihood of getting restful sleep. In the fifth model, high conscientiousness, p < .001, and agreeableness, p = .005, were associated with increased alcohol-related harm reduction behaviors, whereas high extraversion was associated with decreased harm reduction behaviors, p < .001. In the sixth model, extraversion was associated with decreased condom use, p < .001. In the final model, increased levels of conscientiousness, p < .001, and openness, p = .001, were associated with increased consumption of fruits and vegetables per day. In the next set of 4 regression models, risky health behaviors, including cigarette use, alcohol use, binge-drinking frequency, and number of sexual partners, were entered as dependent variables, respectively (see Table 2). In the first model, extraversion, p < .001, and neuroticism, p = .003, were positively associated with cigarette use, whereas conscientiousness was inversely associated with cigarette use, p = .001. In the second model, extraversion was associated with increased alcohol use, p < .001, and conscientiousness was associated with decreased alcohol use, p < .001. In the third model, conscientiousness, p < .001, and agreeableness, p = .002, were inversely associated with binge drinking in the last 2 weeks, whereas extraversion, p < .001, was positively associated with binge drinking. In the final model, high extraversion, p < .001, and low agreeableness, p < .001, were associated with increased number of sexual partners. Although not the primary focus of the present study, it is noteworthy that the block of control variables accounted for a significant amount of variance in most, but not all, of the regression models. As seen in Table 1, a few significant associations were observed among the control variables entered in the first block of predictors and preventive health

behaviors. Specifically, members of Greek organizations were less likely to engage in alcohol-related harm reduction, p < .001; males were less likely to use seat belts, p < .005, and engage in alcohol-related harm reduction, p < .005; and intercollegiate athletes were more likely to engage in moderate/vigorous exercise, p < .001, strengthening exercise, p < .001, and the consumption of more servings of fruits and vegetables p < .01. Table 2 documents how Greek affiliation robustly predicted risky health behaviors, with Greek members significantly more likely to smoke cigarettes, p < .001, consume alcohol, p < .001, binge drink, p < .001, and have a higher number of sexual partners, p < .001. However, with the exception of males reporting higher levels of alcohol use, p < .005, and binge drinking, p < .001, sex and athletic status were weak predictors of risky behaviors. COMMENT This study replicates and extends previous research by examining associations among broad personality traits and several important health behaviors in a sample of college students. Although earlier studies examined personality traits and health behaviors individually, this is the first study conducted with US college students that examines the complete FFM of personality and a wide range of preventive and risky health behaviors. As with results from recent national studies,2 participants in the present sample failed to achieve recommended levels for several goals delineated in Healthy Campus 2010 guidelines3 regarding physical activity, fruit/vegetable consumption, sleep, alcohol use, cigarette smoking, seat belt use, and safe sexual behavior. When sex, intercollegiate athlete status, and Greek affiliation were statistically controlled, results for the

TABLE 2. Hierarchical Linear Regressions (Standardized ) of the Five-Factor Model of Personality Variables on Risky Health Behaviors Cigarette use (last 30 days) Alcohol use (last 30 days) Binge drinking Number of sexual partners (past 2 weeks) (last school year)

Variable

Step 1 Sex Greek status Athletic status R2 Step 2 Openness Conscientiousness Extraversion Agreeableness Neuroticism R2 Adj R2 F

.07 .17* .08 .043* .07 .15* .18* .05 .14* .073* .104 9.44*

.14* .29* .03 .092* .04 .19* .25* .05 .01 .091* .171 16.01*

.24* .31* .08 .136* .02 .18* .29* .13* .01 .120* .246 24.79*

.10 .24* .02 .063* .06 .08 .26* .16* .05 .096* .148 13.62*

Note. For each F, dfs = 8, 574. a For sex, female (n = 428) was coded as 1; male (n = 149) was coded as 2. b Greek affiliation was coded as 0 for no (n = 488) and 1 for yes (n = 95). c Athletic status was coded as 0 for no (n = 466) and 1 for yes (n = 117). * p < .01, 2-tailed.

VOL 58, JULY/AUGUST 2009

77

Raynor & Levine

most part corroborated hypothesized personality-health behavior relations. Conscientiousness was most consistently associated with a health-promoting lifestyle. Relative to those with low levels of conscientiousness, those high in conscientiousness were more likely to wear seat belts, engage in moderate/vigorous strengthening exercise, get enough sleep to feel rested, utilize alcoholrelated harm reduction strategies, and consume fruits and vegetables. Moreover, highly conscientious individuals were less likely to smoke cigarettes, consume alcohol, and binge drink. Although any explanations are admittedly speculative, there are several potential mechanisms linking conscientiousness with health-promoting behaviors. Conscientious individuals may be highly socialized to follow rules and regulations, such as wearing seat belts and not consuming alcohol before reaching the minimum legal age.38 Conscientious individuals tendency to be thorough in decision making may lead to careful consideration of the costs and benefits of engaging in various health behaviors, such as smoking and binge drinking. Finally, conscientious individuals may be successful at delaying immediate gratification while concurrently planning and engaging in preventive behaviors necessary to achieve long-term health goals.39 For example, highly conscientious individuals may adopt regular exercise and consume fruits and vegetables to minimize future risk of developing cardiovascular disease and cancer. The observed associations among extraversion and health behaviors partially supported our predictions. As expected, individuals high in extraversion were more likely to engage in strengthening exercise. However, those high in extraversion did not report engaging in increased moderate or vigorous exercise, which is inconsistent with findings from several studies.7 Results also indicated that individuals high in extraversion were more likely to engage in several deleterious health behaviors, including increased cigarette smoking, increased alcohol use, increased binge drinking, decreased use of alcohol-related harm reduction strategies, increased number of sexual partners, decreased condom use, and decreased likelihood of getting enough sleep to feel rested. In general, the preponderance of the evidence is consistent with prior research demonstrating that individuals high in extraversion engage in a variety of risky behaviors.9,1520 Personality theorist Hans J. Eysenck40 posited that highly extraverted individuals experience chronic cortical underarousal, and as a result, pursue highly stimulating situations to optimize their arousal level. One may speculate that extraverted individuals spontaneously engage in risky health behaviors to fulfill their biologically based needs for excitement, activity, and sensation. We did not make a priori predictions for associations among neuroticism, agreeableness, openness, and health behaviors because of a lack of compelling conceptual rationale or minimal supporting empirical evidence in the college health literature. That said, results indicated that neuroticism and agreeableness predicted a few health
78

behaviors. Relative to those low in neuroticism, individuals high in neuroticism smoked more cigarettes and were less likely to get enough sleep to feel rested. Although these 2 results replicated findings from previous studies,8,18,22 no other significant associations were observed between neuroticism and any of the other health behaviors. Relative to those low in agreeableness, individuals high in agreeableness reported decreased binge drinking, increased use of alcohol-related harm reduction strategies, and decreased number of sexual partners, partially replicating prior research showing that individuals high in agreeableness are less likely to experience alcohol-related disorders.12,13 Another result was that openness predicted increased consumption of fruits and vegetables. Although this finding is novel to the college health literature, it replicated findings from a community sample of adults in which openness predicted several indices of healthy eating.41 Individuals high in openness may be more familiar with and amenable to unconventional dietary practices, which, in the United States, include consuming fruits and vegetables. However, given the lack of other significant relations involving openness, it is reasonable to conclude that this personality dimension plays a minimal role in most health behaviors among college students. Results showed that Greek affiliation was a particularly robust predictor of risky health behaviors. Specifically, Greek members were more likely to smoke cigarettes, consume alcohol, binge drink, and have a higher number of sexual partners, and were less likely to engage in alcoholrelated harm reduction. These findings are highly consistent with previous research indicating that Greek members alcohol and substance use is significantly higher than that of non-Greek members.29,4245 The finding of increased risky sexual behavior is relatively novel, as only 1 previous study documented that Greek members reported more sexual partners in the past 3 months and past year than did non-Greek members.45 Two possible explanations for why Greek members may engage in risky health behaviors are that (1) late adolescents at risk for, or already engaging in, risky behaviors self-select into peer groups and environments that share their preferences, and (2) socialization processes within fraternity and sorority milieus promote engagement in risky behaviors.44 Limitations Several limitations in the current study should be addressed in future research. Although conducting survey research via the Internet is thought to have many positive characteristics (eg, inexpensiveness, access to larger and more diverse participant pools, elimination of data entry) and preliminary evidence suggests that Internet-based findings are consistent with findings derived from traditional methods, potential weaknesses (eg, possible nonserious responses, lack of controlled reporting conditions) also exist.46 Until additional data have validated this relatively new modality, caution should be exercised when interpreting these Internet-based findings.
JOURNAL OF AMERICAN COLLEGE HEALTH

Personality and Health Behaviors

Another limitation is the studys correlational, crosssectional design. As such, it is not possible to infer that personality traits directly caused variability in health behaviors. A viable alternative explanation is that students personality traits affected their choice of situations or contexts, which in turn influenced their health behaviors. For instance, highly extraverted individuals may choose to join Greek organizations and subsequently live in fraternities or sororities. This environment, in turn, may directly promote a range of unhealthy behaviors. We accounted for this specific possibility by statistically controlling for Greek status in all regression analyses. Future studies utilizing longitudinal designs would facilitate tracking students personality and health behaviors as they transition from high school to college and beyond. This would enable researchers to test causal hypotheses regarding direct and indirect relations among personality, environment, and subsequent health behaviors. The generalizability of the present findings may be limited. In comparison to the broader population of American college students, the present sample was relatively homogenous and restricted. Participants were primarily women, predominantly white, and attending a small public liberal arts college in the northeastern United States. Also, compensation for participating involved a relatively small chance to win a $25 gift certificate to the campus bookstore. It is unclear if the present findings generalize to students who were not motivated to volunteer by this compensation. Moreover, it is conceivable that those students who selfselected to participate in this study were more interested in their health and well-being than were those who chose not to participate. If so, a restricted range of variability in health behaviors, and possibly personality traits, may have attenuated hypothesized associations among these variables. Nonetheless, several significant associations were observed, so it is likely that even stronger associations would be detected in a more heterogeneous and comprehensive sample. Future research should attempt replication with a broader range of sample characteristics, including diverse ages, ethnicities, types of higher education institutions, geographic locations, and more male students. Another limitation was the global level of personality assessment. The Big Five are recognized as superordinate domains at the highest level of a hierarchical taxonomy of personality structure. Given that the entire spectrum of personality is condensed to these 5 factors, each of these traits must be broadly operationalized and measured. However, leading contemporary conceptualizations of the FFM incorporate facets, or lower-level traits, that are subsumed within the broader dimensions.4 A lower-level analysis of personality facets may contribute to an improved understanding of health behavior beyond that attainable solely by the Big Five domains. In fact, prior research using such a fine-grained approach identified relations between lower-order personality facets and health behaviors that were different from those noted at the domain level.13 Thus, future research should incorporate a facet-level analysis in addition to the domain level to increase accuracy in predicting individual differences in health behaviors.
VOL 58, JULY/AUGUST 2009

Future Research The present findings have several implications for health promotion interventions aimed at individual students and campus communities. One approach would be to identify those individuals with personality characteristics associated with preventive or risky health behaviors and to intervene with techniques that specifically target these individual risk factors. Along these lines, Conrad and her colleagues4749 have developed and tested cognitive-behavioral interventions designed to target a few key personality dimensions that are consistent predictors of risky health behaviors, particularly alcohol and substance misuse. For instance, in each of Conrad et als randomized controlled trials,4749 individuals self-reporting high levels of sensation seeking, a facet of extraversion, were targeted because of the consistently strong association between this trait and heavy alcohol-use and alcohol use disorders among adolescents and adults.50 Although procedures varied slightly across studies, common features of the sensation-seeking matched intervention included psychoeducational information on this personality dimension, as well as personality-specific cognitive-behavioral exercises designed to facilitate more adaptive coping strategies. Results from these studies indicate that the sensation-seeking intervention was associated with significantly decreased levels of binge drinking among participants who initially self-reported high levels of sensation seeking at baseline.4749 It is important that 1 of the studies demonstrated that the sensation-seeking matched intervention was more effective than an attentional control intervention and partially supported the advantage of a personality-matched intervention over a personalitymismatched intervention.47 Findings of the present study suggest that low-conscientious individuals comprise another subset of college students at high risk for an unhealthy lifestyle. For instance, given that low-conscientious individuals may lack the tendency to strive for competence and achievement, a promising approach for these at-risk individuals may be motivational enhancement. A recent meta-analysis of individual-level interventions to reduce college student drinking indicated that individual, face-to-face motivational interviewing was highly effective in reducing alcohol-related problems.51 Another aspect of low-conscientious individuals is the tendency to lack self-discipline and deliberativeness and to seek immediate rather than delayed gratification. Conrod et al49 developed an intervention targeting impulsiveness and demonstrated that this personality-matched intervention could significantly reduce risky behavior (ie, shoplifting) over time among British adolescents. Moreover, a variety of other conscientiousness-related deficits, such as the tendency to be disorganized and unreliable, can be targeted with a variety of structure-enhancing strategies, including regular appointments, self-monitoring, directly observing health behaviors, tailoring regimens to daily habits, and appointment and behavior change reminders via phone and e-mail. However, with the exception of the aforementioned
79

Raynor & Levine

interventions by Conrod et al4749 targeting the traits of sensation seeking and impulsiveness, the effects of other personality-based interventions have not been tested and warrant future study. Finally, although brief personality-specific psychotherapeutic interventions for those with risky personality types hold much promise, it is likely that some at-risk individuals would not participate in such preventive measures. As such, college administrations should complement the individualbased level of personality-based intervention with health promotion interventions aimed at the campus community. For instance, convenient access to a variety of less risky, alternative social activities that are highly active and stimulating (eg, rock climbing, X-Game competitions, studentled emergency response services) would provide healthy alternatives for highly extraverted individuals.52 Although this type of public health intervention apparently has not been empirically tested in a college setting, the provision of health promotion based on individualized reasons for failing to adopt health behaviors may be particularly efficacious in utilizing limited resources on a campus-wide scale.
ACKNOWLEDGMENT

This study was funded in part by an Incentive Grant from the Geneseo Foundation to Dr Raynor and by a State University of New York Alcohol & Other Drug Grant to Dr Levine.
NOTE

For comments and futher information, address correspondence to Dr Douglas A. Raynor, State University of New York, Department of Psychology, 1 College Circle, Geneseo, NY 14454, USA (e-mail: raynor@geneseo.edu).
REFERENCES 1. US Department of Health and Human Services. Healthy People 2010: Understanding and Improving Health. 2nd ed. Washington, DC: US Government Printing Office; 2000:751. 2. American College Health Association. American College Health AssociationNational College Health Assessment spring 2006 reference group data report (abridged). J Am Coll Health. 2007;55:195206. 3. American College Health Association. Healthy Campus 2010: Making It Happen. Baltimore, MD: American College Health Association, 2002:4446, 5455, 6873, 7882, 94100. 4. John OP, Srivastava S. The Big Five: history, measurement, and development. In: Pervin LA, John OP, eds. Handbook of Personality: Theory and Research. New York, NY: Guilford Press; 1999:102138. 5. Costa PT, McCrae RR. The NEO Personality Inventory-R: Professional Manual. Odessa, FL: Psychological Assessment Resources; 1992:3955. 6. McCrae RR, Costa PT. Personality trait structure as a human universal. Am Psychol. 1997;52:509516. 7. Rhodes RE, Smith NEI. Personality correlates of physical activity: a review and meta-analysis. Br J Sports Med. 2006;40:958965. 8. Gray EK, Watson D. General and specific traits of personality and their relation to sleep and academic performance. J Pers. 2002;70:177206. 9. Ingledew DK, Ferguson E. Personality and riskier sexu80

al behaviour: motivational mediators. Psychology and Health. 2007;22:291315. 10. Arthur W Jr, Graziano WG. The Five-Factor Model, conscientiousness and driving accident involvement. J Pers. 1996;64:593618. 11. Arthur W Jr, Doverspike D. Predicting motor vehicle crash involvement from a personality measure and a driving knowledge test. J Prev Interv Community. 2001;22:3542. 12. Martin ED, Sher KJ. Family history of alcoholism, alcohol use disorders and the Five-Factor Model of personality. J Stud Alcohol. 1994;55:8190. 13. Ruiz MA, Pincus AL, Dickinson KA. NEO PI-R predictors of alcohol use and alcohol-related problems. J Pers Assess. 2003; 81:226236. 14. Kashdan TB, Vetter CJ, Collins L. Substance use in young adults: associations with personality and gender. Addict Behav. 2005;30:259269. 15. Allsopp JF. Personality as a determinant of beer and cider consumption among young men. Pers Individ Dif. 1986;7:341 347. 16. Martsh CT, Miller WR. Extraversion predicts heavy drinking in college students. Pers Individ Dif. 1997;23:153155. 17. Benjamin L, Wulfert E. Dispositional correlates of addictive behaviors in college women: binge eating and heavy drinking. Eat Behav. 2005;6:197209. 18. Spielberger CD, Jacobs GA. Personality and smoking behavior. J Pers Assess. 1982;46:396403. 19. Dahlen ER, White RP. The Big Five factors, sensation seeking, and driving anger in the prediction of unsafe driving. Pers Individ Dif. 2006;41:903915. 20. Schwebel DC, Severson J, Ball KK, Rizzo M. Individual difference factors in risky driving: the roles of anger/hostility, conscientiousness, and sensation-seeking. Accid Anal Prev. 2006;38:801810. 21. Buckworth J, Granello DH, Belmore J. Incorporating personality assessment into counseling to help college students adopt and maintain exercise behaviors. J Coll Counsel. 2002;5:1525. 22. McChargue DE, Cohen LM, Cook JW. The influence of personality and affect on nicotine dependence among male college students. Nicotine Tob Res. 2004;6:287294. 23. Ham LS, Hope DA. College students and problematic drinking: a review of the literature. Clin Psychol Rev. 2003;23:719 759. 24. Courneya KS, Hellsten LM. Personality correlates of exercise behavior, motives, barriers and preferences: an application of the Five-Factor Model. Pers Individ Dif. 1998;24:625633. 25. Fine LJ, Philogene S, Gramling R, Coups EJ, Sinha S. Prevalence of multiple chronic disease risk factors: 2001 national health interview survey. Am J Prev Med. 2004;27:1824. 26. Lemos-Girldez S, Fidalgo-Aliste AM. Personality dispositions and health-related habits and attitudes: a cross-sectional study. Eur J Pers. 1997;11:197209. 27. Ingledew DK, Brunning S. Personality, preventive health behaviour and comparative optimism about health problems. J Health Psychol. 1999;4:193208. 28. Vollrath M, Torgerson S. Who takes health risks? A probe into 8 personality types. Pers Individ Dif. 2002;32:11851197. 29. Turrisi R, Mallett KA, Mastroleo NR, Larimer ME. Heavy drinking in college students: who is at risk and what is being done about it? J Gen Psychol. 2006;133:401420. 30. Irwin JD. Prevalence of university students sufficient physical activity: a systematic review. Percept Mot Skills. 2004;98:927 943. 31. American College Health Association. National College Health Assessment: Reliability and Validity Analyses. Baltimore, MD: American College Health Association; 2004:234. 32. Gow AJ, Whiteman MC, Pattie A, Deary IJ. Goldbergs IPIP Big-Five factor markers: internal consistency and conJOURNAL OF AMERICAN COLLEGE HEALTH

Personality and Health Behaviors current validation in Scotland. Pers Individ Dif. 2005;39:317 329. 33. Milas G, Mlacic B, Kratohvil A. Adolescent personality and self-esteem: an analysis of self-reports and parental-ratings. Soc Res J for Gen Soc Issues. 2007;16:213236. 34. Mlacic B, Goldberg LR. An analysis of a cross-cultural personality inventory: the IPIP Big-Five factor markers in Croatia. J Pers Assess. 2007;88:168177. 35. Campbell DT, Fiske DW. Convergent and discriminant validation by the multitrait-multimethod matrix. Psychol Bull. 1959;56:81105. 36. Buchanan T, Johnson JA, Goldberg LR. Implementing a 5factor personality inventory for use on the internet. Eur J Psychol Assess. 2005;21:115127. 37. Nettle D. Psychological profiles of professional actors. Pers Individ Dif. 2006;40:375383. 38. Roberts BW, Walton KE, Bogg T. Conscientiousness and health across the life course. Rev Gen Psychol. 2005;9:156168. 39. Booth-Kewley S, Vickers RR. Associations between major domains of personality and health behavior. J Pers. 1994;62:281 298. 40. Eysenck HJ. Biological dimensions of personality. In: Pervin LA, ed. Handbook of Personality: Theory and Research. New York, NY: Guilford Press; 1990:244276. 41. Goldberg LR, Stryker LA. Personality traits and eating habits: the assessment of food preferences in a large community sample. Pers Individ Dif. 2002;32:4965. 42. Wechsler H, Kuh G, Davenport AE. Fraternities, sororities and binge drinking: results from a national study of American colleges. NASPA J. 1996;33:260279. 43. Cashin JR, Presley CA, Meilman PW. Alcohol use in the Greek system: follow the leader? J Stud Alcohol. 1998;59:6370. 44. McCabe SE, Schulenberg JE, Johnston LD, OMalley PM, Bachman JG, Kloska DD. Selection and socialization effects of fraternities and sororities on US college student substance use: a multicohort national longitudinal study. Addiction. 2005;100:512524. 45. Scott-Sheldon LAJ, Carey KB, Carey MP. Health behavior and college students: does Greek affiliation matter? J Behav Med. 2008;31:6170. 46. Gosling SD, Vazire S, Srivastava S, John OP. Should we trust Web-based studies? A comparative analysis of 6 preconceptions about Internet questionnaires. Am Psychol. 2004;59:93104. 47. Conrad PJ, Stewart SH, Phil RO, Cote S, Fontaine V, Dongier M. Efficacy of brief coping skills interventions that match different personality profiles of female substance abusers. Psychol Addict Behav. 2000;14:231242. 48. Conrad PJ, Stewart SH, Comeau N, Maclean AM. Efficacy of cognitive-behavioral interventions targeting personality risk factors for youth alcohol misuse. J Clin Child Adolesc Psychol. 2006;35:550563. 49. Conrad PJ, Castellanos N, Mackie C. Personality-targeted interventions delay the growth of adolescent drinking and binge drinking. J Child Psychol Psychiatry. 2008;49:181190. 50. Sher KJ, Bartholow BD, Wood MD. Personality and substance use disorders: a prospective study. J Consult Clin Psychol. 2000;68:818829. 51. Carey KB, Scott-Sheldon LAJ, Carey MP, DeMartini KS. Individual-level interventions to reduce college student drinking: a meta-analytic review. Addict Behav. 2007;32:24692494. 52. Roberti JW. A review of behavioral and biological correlates of sensation seeking. J Res Pers. 2004;38:256279.

VOL 58, JULY/AUGUST 2009

81

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.