Did an AIDS Peer Education Program Change... Students' Behaviors?

Abstract. Students who have attended the AIDS peer education program (APEP) at Florida Atlantic University have consistently evaluated it in an overwhelmingly positive manner. This has inspired the university staff to move on to another issue of concern: Is attendance at such a peer education program associated with HIV-related changes in behavior? To answer this question, the authors surveyed a random sample of first-year students by mail at the beginning and end of the 1991/92 school year. In the interim, some students participated in an APEP and some did not - resulting in a before-after design. Those who attended an APEP reported they were more likely than nonattendees to engage in behaviors that were aimed at preventing HIV infection. The validity problems inherent in self-selecting attendance and in surveying by mail, especially in regard to causation versus association, lead us to attempt a face-to-face interview design in future evaluations. With stricter control over other potential intervening variables, we may be more likely to obtain data that relate change in behavior to the APEP.
Key Words: AIDS, behavior change, peer education, sexual behavior, university and college students

In a previous issue of The Journal of American College Health, we and our colleague wrote about a then innovative classroom AIDS Peer Education Program (APEP) instituted in 1984 by the nursing staff of the student health service (SHS) at Florida Atlantic University (FAU) (1). We described a classroom-based program in which volunteer student peer educators were trained to enter regularly scheduled classrooms (with faculty permission) to make presentations on the causes, prevention, and treatment of human immunodeficiency (HIV) disease. In their classroom appearances, the students were accompanied by a registered nurse and, from time to time, by an individual who was HIV-positive. Immediately afterward, all students completed an anonymous evaluation of the presentation before they left the room. Over the years, several thousand of these evaluations have been reviewed by the SHS, and we have found that more than 90% of the responses have been positive. Critical comments or suggestions for improvement have been given careful attention, and the APEP has now been revised - usually by adding new topics of particular interest to the students. In this article, we will consider our attempt to go beyond measuring student satisfaction with the program and will move toward the more elusive goal of measuring how HIV-related behavior changed as a result of exposure to the APEP.

In our follow-up study, we decided to test the hypothesis that experiencing the APEP would result in measurable behavior change in student's HIV-related behaviors. We also decided to focus on the youngest population available to us - first-year college students. The focus on first-year students was prompted by increasing concern over the early age at which many individuals are becoming infected with HIV. The Centers for Disease Control and Prevention (CDC) has reported that more than one in five cases of AIDS is diagnosed in persons from 20 to 29 years of age. Because the average time between infection with HIV and development of advanced HIV disease is about 10 years, it is clear that many of those diagnosed in their twenties were advanced disease became infected when they were teenagers (2,3). Complications associated with advanced HIV disease are now the sixth leading cause of death for persons aged 15 to 24 years (4-6). High school and college students' high-risk behaviors, therefore, are of considerable concern among those who want to encourage changes in behavior.

For our before-after study, we sent an anonymous mail survey to a random sample of entering first-year students at the beginning of the school year and a follow-up survey at the year's end. We hypothesized that some of these students would attend an APEP in their regularly scheduled classes during the interim and that others would not - creating a design with both an experimental group and a control group.

The university registrar's office used a computer program to provide us with a random selection of the names of 200 incoming first-year students and their home addresses at the start of the 1991/92 school year. Financial constraints dictated the size of the sample. In August 1991, the 200 randomly selected students were sent Survey 1, with an explanatory cover letter and a postage-paid return envelope. The anonymous questionnaire elicited attitudes toward HIV and HIV-related behaviors. Sixty-eight incoming students, or 34% of those who received the mailing, responded to the first survey, which contained 28 questions. In May 1992, a follow-up survey was sent to the original 200 persons whose names had been generated by the computer. They were invited to respond to Survey 2, which contained 29 questions, and 43 students did so. We used an anonymous code on both the August and May mailings so that we could identify which surveys were completed both times by the same respondent and then compare individual responses at the beginning of the school year with those at the end of the year. The additional question on Survey 2 asked whether or not the respondent has attended an APEP during the academic year. The survey instrument has been pretested on a first-year English class of 20 students, and individual items had been pretested several times when they were used in other, related SHS research. It was also reviewd and approved by the Human Subjects Board, the university's institutional review board.

When only 68 persons responded to the first survey, we considered sending a follow-up reminded to increase the size of the sample. That action, however, could have interfered with the research design, in that the follow-up letter could have arrived after the start of the school year and possibly after some of the students had already been exposed to the APEP in class. Because we wanted all responses to Survey 1 to be collected before involvement with the APEP was possible, we did not send follow up reminders. In addition, attrition may have reduced the sampling frame because some of the 200 incoming first-year students may never have arrived for the fall semester, and some may not have survived their first year of college. Although we considered checking with the registrar's office for the names and addresses of those who failed to arrive the following May, we decided against that action. Our goal of preserving the anonymity of respondents, in view of the intrusive nature of the survey, could have been compromised by any attempt to reduce the original list of 200.

At the end of the process, we identified a final sample of 24 individuals who had answered both surveys; 11 of the 24 had attended an APEP during the school year, 13 had not done so. This article deals only with the responses of the 24 students who provided answers on both surveys. The small sample size greatly limits the usefulness of our data, but the design could be expanded and replicated.

We made several comparisons between the two groups (APEP = 11 and non APEP = 13). Because the resulting subsamples were small, sophisticated tests of statistical significance were inappropriate. Thus, we treated the data as qualitative, with no attempt to claim causal relationships - we sought only to discover whether there was an association. We are addressing only this sample and are not generalizing to all first-year college students in view of the limitations in external validity. We obtained two types of responses. The first dealt with actual reported behavior change, and the second dealt with reported likelihood of adopting safer sexual behavior.

Some of the survey questions asked about the likelihood of adopting risk reducing behavior. Those who attended an APEP during their first year of college reported that they were more likely than those who did not to

We also made comparisons between students' reported intentions at the beginning of the first year of college and at the end of the year. Those who attended an APEP during their first year of college reported that during the school year they were more likely to engage in risk-reduction behavior (even though they had not done so at the start of the school year) than those who had not been exposed to the APEP. The activities they reported were likely to engage in were to

We experienced difficulty in obtaining enough responses to our two 200 survey mailings with the result that we received only 24 usable responses for the final analysis; this limited our use of quantitative statistical analysis. The research design did not have adequate built-in controls to allow us to demonstrate that exposure to the APEP was the cause of self reported behavior change. The discussion of the differences between APEP-exposed students and others have been couched in terms of self reported behavior changes and intention of changing certain HIV-related behaviors between the start of the first year of college and the end of that year. Nonetheless, the qualitative data support further investigation into a causal effect, especially because definite behavioral changes were reported during the school year in the areas of HIV-antibody resting and condom use.

External validity could have been improved if the sample had been considerably larger. Internal validity could have been improved if greater control had been built into the design - such as the measurement of other variables that might be associated with behavior change (eg, knowledge of a friend, relative or acquaintance become HIV-infected; exposure to HIV educational material outside the APEP; the length of time that had passes between attending the APEP and completing the second survey; concern over one's current health status).

Upon reflection, we question whether the kind of in-depth, behavioral change information we are interested in can be obtained effectively through the rather remote process of a mail survey. Although a recent report on a classroom survey approach indicated that large samples and strict controls are possible (7), we are moving toward a face-to-face interview design. Therefore, we have decided that our future research in this area will be based on face-to-face interviews - a design that will allow for more complex, multifaceted questioning of respondents. Researchers will be limited to asking about specific behaviors of direct interest to their study, but can also ask about other areas and behaviors in the students' lives that may be affecting their perception of HIV. A mail survey that tries to record or control for the many potential intervening variables (beside the APEP) that could have a behavioral effect would probably not be tolerated or answered by most recipients because of its length. In a "captive" interview situation, we hope the respondent would be somewhat more available for the kind of detailed proving we desire.

In addition to exploring the use of the interview approach, we intend to explore some of the models of behavioral change discussed in the literature that focus upon the phenomenon of change as a process occurring in stages over time (8-10). We are confident that this combination of approaches will yield more valid data that is capable of eventually allowing us to go beyond association and in the direction of causation (to the extent that this is possible in dealing with complex behavioral phenomena).

These data, although qualitative because of the small size of the sample, indicate that these students are in the "contemplation stage" of the process of behavioral change that was recently discussed by Prochaska et al (11). Health educators can design programs with that in mind as they attempt to move students toward the next stage - the "action stage." We have been aware of the overwhelmingly positive response the APEP has received from participants since its inception in 1987. This new study provides some indication that classroom exposure to the program is associated with increase HIV-antibody testing and condom use, as well as stated intentions of reducing HIV-related behaviors. These useful data will allow us to plan a future, in-depth interview study to prove the extent to which intentions are translated into actions and the process by which that occurs.

2. Centers for Disease Control. Preventing risk behaviors among students. HIV/AIDS Prevention Newsletter. October 1992;3:1-2.

3. Centers for Disease Control. AIDS cases by sex, age at diagnosis, and race/ethnicity, reported through December 1992, United States. HIV/AIDS Surveillance. February 1993;4:14.

4. Centers for Disease Control. High school students may be changing high risk sexual behavior. HIV/AIDS Prevention Newsletter. December 1992;3:14.

5. Haffner D. Youth still at risk, yet barriers to education remain. SIECUS Report. October/November 1992:10-12.

6. Hein K. Getting real about HIV in adolescents. Am J Pub Health. 1993;83:492-494.

7. Turner J Korpita E, Mohn L, Hill W. Reduction in sexual risk behaviors among college students following a comprehensive education intervention. J Am Coll Health. 1993;41:187-193.

8. Silven D. Behavioral theories and relapse. Focus: A Guide to AIDS Research and Counseling. January 1993;8:1-4.

9. Prochaska J, DiClemente C. Trans-theoretical therapy: Toward a more integrated model of change. Psychotherapy Theory, Research and Practice. 1982;19:276-288.

10. Prochaska J, DiClemente C. Stages and processes of self-change of smoking: Toward an integrative model of change. J Consult Clin Psychology. 1983;51:290-295.

11. Prochaska J, DiClemente C, Norcross J. In search of how people change: Applications to addictive behavior. Am Psychologist. 1992;47:1102-1114.

Did an AIDS Peer Education Program Change First-Year College Students' Behaviors?