Poor women, minority women, and homeless women are at high risk for HIV infection. To help, a relevant AIDS prevention educational program must interest its target population, exhibit models who are perceived as demographically and contextually similar, and influence behavior change. Yet for the design of such programs, there is little evidence as to the specific demographic (race) and contextual (behavior and situation) variables that make for effectiveness. The objective of this cross-sectional, group-comparison descriptive study is to delineate the relevant demographic and contextual variables among poor and precariously housed African-American, Caucasian, and Hispanic women. The first specific aim is to investigate systematically the three groups' perceptions of the behaviors and situations that increase women's chance of contracting AIDS, and on that basis to develop AIDS prevention video vignettes.The second specific aim is to compare the three groups' evaluation of the vignettes in terms of 1) which vignettes they pay attention to and which interest them, 2) which vignettes they identify with racially as well as behaviorally and situationally, and 3) which vignettes they perceive as most likely to change their behavior. To accomplish the first aim, separate focus groups for each of the three racial groups will be conducted to identify AIDS risk behaviors and situations. The hypothesis is that the three will not differ in their reported AIDS risk behaviors and situations. Based on these data, two sets of vignettes will be developed, according to the principles of Bandura's social cognitive learning theory. In the first set of vignettes, race will be systematically varied, with behavior and situation held constant across racial groups. In the second set, behavior and situation will be systematically varied, with race held constant.
The second aim will be accomplished by conducting separate focus groups for each of the groups where self-administered rating scales will be rated individually after the women have viewed the vignettes followed by discussion of focus group questions. The hypotheses are that 1) there is no difference among the three groups' interest in and attentiveness to the vignettes, 2) there is no difference among the groups' perceptions of the vignette to which they identify as being racially, and behaviorally and situationally similar to themselves, and 3) there is no difference among the groups' perception of the vignette most likely to promote behavior change.
