The NCI-IDD team worked with the state of Texas to pilot test audio telephone administration of the Child Family Survey.

Through examination of the surveyor and respondent feedback data, we see that respondents
rated the experience of doing the survey by phone as relatively easy. However, qualitative
information indicated numerous barriers, challenges, and frustrations related to the phone mode
such as distraction, communication difficulties, length of survey questions, and the length of the
survey itself. There were also differences in completion rate, with mail/direct entry respondents
having significantly fewer questions with missing data in the later portions of the survey,
potentially indicating that phone respondents were more likely to end their surveys before they
were complete.

Results of the pilot study of the phone mode of administration of the Child Family Survey indicate
that there were significant differences between the personal and demographic characteristics of
respondents who answered by phone, and those that responded by mail/direct entry. Although
we are unsure the impact of these differences on the representativeness of the data, we believe
that these differences merit further examination.

Furthermore, there were significant mode differences in the actual responses to the survey
questions and the rates of missing data and “don’t know” responses. These differences
encourage us to ask whether the survey, as it is written, can translate well from being
administered visually to auditorily.

Based on this study, we do not believe that NCI-IDD should proceed with administration of the
family survey via phone mode until further testing and research is done.