Survey Design for TANF Caseload Project: Summary Report and Recommendations. Common Measures Used, Pros and Cons


Some surveys (such as Alameda, Missouri, and Illinois) used a detailed grid format to collect data on household composition. For each member in the respondents household, these surveys gathered the name, date of birth or age, and relationship of that person to the respondent. The Alameda survey included additional questions on whether the person lived with the respondent during the prior year and, if not, where the person was living before, and whether the respondent supported the person financially during that time. The Missouri survey included an extra question on whether the person lived with the respondent during the prior month but asked no follow-up questions. The Illinois survey included additional questions on whether the person living in the household had a job, whether the person was a minor, and, if so, whether the minors other parent lived in the household.

The Nebraska survey collected information on household composition in a different format. One question asked for the total number of people living with the respondent. Two more questions asked for the number of children living in the household, grouped by age (0 to 5 and 6 to 17 years). No questions collected the relationship of the household members to the respondent. The Iowa Child Impact Survey did not collect any detailed household information because the information was already collected in a prior survey (the Iowa Core Survey), using the detailed grid format described above. The detailed grid format for collecting household composition provides more precise information about each household member. However, the method is time-consuming and tends to produce a higher level of item nonresponse in the form of refusals than is achieved using anonymous methods. We believe that collecting the information in a more general way, such as in the Nebraska survey, will obtain the requisite data on the number of household members and the age ranges of any children. However, it will not provide information about the relationship of household members to the respondent.

A few surveys collected information on respondents children who were not living with the respondents. The Nebraska, Alameda, and Illinois surveys asked whether the respondent had any children under age 18 who did not live with them and, if so, how many. The Illinois survey also asked where the children were living and why they were living there. The WES asked only for the number of children under age 18 living outside the respondents household. We believe that it is important to gather the number of children under age 18 the respondent has who live outside the respondents household, and also where they are living, and why.

View full report


"report.pdf" (pdf, 367.08Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®