The Center for Survey Methods Research conducted cognitive testing of proposed new questions on costs of child care to be included in the 2001 SIPP Work Related Expenses Module. In conducting the testing, we used the entire module including questions on both child care and commuting costs. This report documents the results of this cognitive pretesting. Interviews were conducted with working mothers with children under age 15 who paid for child care across the socio-economic spectrum.
The results showed respondents had problems with the transportation series for people who drive to work. The initial question asks whether respondents drive to work, but the response category and following questions assume that they drive their own cars. Respondents who do not drive their own cars were confused by these questions. Depending on the intent of the sponsors, data may be incorrect for respondents who do not report costs because they do not drive their own vehicles. This is a topic that should be reconsidered when the question is revised.
Administration of the child care cost questions showed some problems, and minor changes in question working are recommended. Respondents who use subsidized child care tended not to report it. The major problem with this series was that respondents tended not to report costs in weekly amounts, but rather in whatever frequency they paid the bills. Depending on how interviewers deal with this situation in the field, this may or may not be a problem. The implications of this finding pertain more to interviewer training than to question wording.
