Hard Lives, Mean Streets: Violence in the Lives of Homeless Women

Appendix A: Study Methodology

Appendix A Study Methodology The Florida Four-City Study of Violence in the Lives of Homeless Women was designed to be a multisite, statewide study. As indicated in the text, the literature on violence and homeless women tends to be dominated by single-city or even singleshelter (or single-facility) surveys. We felt that it was important to obtain data from multiple sites in order to avoid overgeneralizing results that might be unique to a particular site. In general, di=erences across the four cities in the study were substantively minor and statistically not signi>cant, with a few important exceptions discussed later in this appendix. So we may be reasonably con>dent that our >ndings generalize at least to urban homeless women in Florida, if not to homeless women everywhere in the United States. Initial Focus Group Altogether, over 700 homeless women participated in the study, development of which began with a focus group in November 2002 at an Orlando domestic violence shelter. This initial focus group involved six women who were part of an outpatient group of domestic violence survivors who had also experienced homelessness . The purpose of this focus group was to pinpoint the key issues relevant to homeless women and their experiences with violence, and to make sure that the larger study not only re?ected the topics that had surfaced in our review of the scienti>c literature, but that it adequately re?ected the life stories of the women we intended to study. Many themes and perspectives emerged from this initial focus group and were used over the next several months to develop a survey questionnaire. The Survey The survey questionnaire was designed to be administered by interviewers at each of the four sites across Florida (Orlando, Tampa, Jacksonville, and Miami, the four largest metropolitan areas in the state). Interviewer training in each of the four sites commenced shortly after the questionnaire was completed, and as soon as training was completed, interviewing began. Interviewers in each site were recruited from 164 Appendix A among existing shelter sta=—mostly case managers, intake workers, and counselors . All our interviewers were highly experienced in dealing with homeless women and their problems, and all took on their interviewing jobs as a supplement to their normal work roles. Our interviewers dealt with issues of violence and victimization on a daily basis. They were already screened, trained, and provided with resources to deal with any adverse personal reactions that respondents might have in the course of the interview. In fact, although some women had strong emotional reactions during the interview (hardly a surprise, given the subject matter), none of these reactions was strong enough to require professional intervention or to rise to the level of an adverse event that had to be reported to an institutional review board. Interviewers were given advice during training about how to deal with respondents ’ signs of emotion or discomfort, including allowing the interviewee to take breaks, demonstrating empathy, discontinuing the interview, and so on. We also reinforced the point that if the interview produced any sign of renewed victimization or traumatization, the interviewee must be referred to the appropriate clinical sta= person at the shelter. However, to our knowledge, no such signs occurred. As part of the training, interviewers were instructed to record nonverbal cues in the margins of the survey instrument. These cues were used as a way to help the interviewer determine whether to probe further, and how to assess the interviewee’s level of discomfort, reluctance, or other emotional issues. Interviewers were trained to conduct their normal intake process, and during intake to ask the client if she was willing to participate in a study conducted by faculty at the University of Central Florida and funded by the National Institute of Justice. The purpose of the interview, clients were told, was to gain insight and understanding into the lives of women who were experiencing homelessness. Prospective participants were also told that they had been chosen at random. All participants signed a consent form that outlined the purpose and goals of the study. In addition, they were given a copy of this form in case they wanted to refer to it later. The form contained contact information for the study director, so that participants with questions could call him. All aspects of the study protocol, including the consent forms, survey questionnaire , guiding questions for the focus group, and the protocols for the in-depth personal interviews, were reviewed and approved by the...