Sampling.Homeless participants were first selected using a convenience sample of people experiencing homelessness in Nashville wherein no ethnicity, gender, adult age, or other demographic factor was used to target or exclude participants. Participants were only excluded from the study if they were under 18, they were cognitively unable to understand the risks and/or the content of the study, or if they did not speak English. Participants were interviewed at shelters, on the street, and in advocacy centers. Once the initial sample was gathered (n ≈ 75), specific demographic markers (gender, age, ethnicity, probable living situation) were targeted using snowball sampling to make the sample more representative of the likely population of people experiencing homelessness in Nashville. The final sample size (n = 105) is based on a predicted level of theoretical saturation (Strauss & Corbin, 1998), wherein new data from a representative sample does not vary results significantly.
Sample demographics gathered in the interviews are as follows: The average age of respondents was 45.33 (median 47) years old; 75% of interviewees were male, 25% were female; 62.16% of participants self-identified as Black/African American, 29.73% self-identified as White/Caucasian, and 8.10% self-identified as other/mixed-ethnicity; 54% were single, 2.7% were married, 21.6% were divorced, and 13.5% were widowed (8% no response); Respondents have lived in Nashville for an average of 35.49 years (median 12 years); 2.7% state that they are veterans of the armed forces; Median income for respondents was $2,800 (mean $4,641.30); 71.4% of participants had been homeless for one year or more prior to the interview; Participants had spent an average of 4.96 years (median 3 years) on the streets or in shelters; 83.78% of respondents live alone (unaccompanied). Overall, 52% of interviewees meet the criteria for chronic homelessness.
Key informants (outreach and service workers, program administrators) were interviewed, to determine service usage and operating costs to estimate the costs of providing services for people experiencing homelessness in Nashville. Based on an initial list from Clifton Harris and known community resources, a snowball sampling procedure was used to determine further service providers. Service costs were sought for addiction treatment, advocacy, case management/referrals, child care and education, clothing, communications (telephone, internet, etc.), counseling, documents, education, emergency care/transport, financial services, food, health care, housing/shelter, incarceration, job training, laundry, legal/courts, mental/behavioral health, pastoral care, personal care items, and transportation. Some costs were located as separate sums while others were part of indivisible budgets. Some costs were not available and had to be estimated based on existing data or excluded from the total.