Sample and Data Collection Design

The initial WES sample consisted of 874 single mothers who were on the welfare rolls in February of 1997. The sample was systematically drawn with equal probability from an ordered list of the universe of active single mother cases of the Michigan Family Independence Agency. The sample was geographically limited to one urban Michigan county. Cases were proportionately selected by zip code, race (White, non-Hispanic or African-American), and age (18-54 years).

The Survey Research Center of the Institute for Social Research at the University of Michigan conducted the data collection effort. SRC interviewers administered face-to-face standardized interviews using a paper-and-pencil survey instrument. Interviews have slightly increased in length over the waves. The average interview length across all five waves was about 75 minutes (ranging from 60 minutes in the first two waves to about 85 minutes in the next three).

Prior to each data collection, interviewers attended a two-day study specific training on the content and administration of the survey instrument, as well as a range of activities such as tracking, addressing respondent concerns, and interviewing on sensitive topics, such as domestic violence and mental health. Overall, the study has maintained a fairly stable group of interviewers who have worked all waves of the study.

Most of the interviews at the first wave of WES were conducted in the Fall of 1997. Waves 2 and 3 were conducted at approximately one-year intervals in the Fall of 1998 and Winter of 1999, respectively. Wave 4 was collected from November 2001 through early 2002; Wave 5, from September 2003 through early 2004.

At each wave, we follow respondents to the previous wave but we do not attempt to recontact non-respondents. The number of completed interviews for each wave and the response rate (calculated as the number of completed cases divided by the number of sample cases) are outlined in the table below: