Predictors of HIV Risk Among College Students: A CHAID Analysis1

Abstract

SUMMARY The purpose of this study was to identify predictors of risk behaviors among college students and to determine how students differed in HIV risk practices. Participants were from six colleges and universities in a large southeastern metropolitan area. The present analysis was limited to participants who were 18 to 25 years of age; single; White, African American, or Asian; and reported initiation of sexual intercourse. The results demonstratedsignificant associations of gender and race with having sex within the past3 months, number of partners, condom use, length of time one knew one's sexual partner, substance use, and asking one's partner about sexualhistory. Using Chi‐Square Automatic Interaction Detection (CHAID), the strongest predictor of condom use was noted to be gender, and the strongest predictor of number of partners was race.The findings of this study indicate that HIV risk factors vary dependingon a college student's ethnicity, gender, academic status, and substanceuse. Most importantly, African American men, particularly men who use drugs, and White and Asian men who use alcohol tend to report higher numbers of sexual partners. Females, particularly upperclass White females, and White men who do not use alcohol report infrequent condom use. These findings areuseful for health educators who have begun to recognize the importance ofconsidering the HIV risk factors of students and incorporating informationrelated to these factors into HIV prevention programs. From the results ofthis study, we have several suggestions that health educators and researchersmay want to consider when developing programs for college students:1. Although students may not need extensive information about basic HIVfacts, they may need reinforcement of the information. In particular, students need to be made aware that lambskin condoms do not offer protection against HIV transmission.2. Females need to be encouraged and taught the skills to negotiate theuse of and to actually use condoms. These skills need to be reinforced throughout college, particularly as women adopt other methods of birth control (i.e., the pill).3. The general acceptance of condom use among African Americans and Asians may be a useful tool for the health educator. Promoting and reinforcingthis norm and behavior, while addressing some of the other risk factors inthese groups, may be a particularly effective approach to HIV prevention.4. For White students whose condom use rates are lower, messages can include the use of condoms for STD and pregnancy protection. Messages for male students could include control over preventing pregnancy or sharing control for pregnancy prevention.5. Couples often abandon using condoms when they feel they know each other “well enough.” Some guidelines for when to give up condoms might also be given to provide couples with responsible and realistic directions for minimizing their risk when choosing to engage in unprotected sex. These guidelines would include HIV testing two times at least 6 months apart, no evidence of STDs in either partner, and a commitment of both partners to a monogamous relationship.6. When addressing the number of sexual partners as a risk factor, African American men and women may hold the greatest potential for change, given the greater likelihood of multiple partners in these groups.7. With respect to drug and alcohol use, health educators may want to include discussions about the influence of drugs (particularly for African American students) and alcohol (particularly for White and Asian students) on inhibiting logical decision‐making as it relates to sexual intercourse.