• Presentation

Does Voice Matter for Youth Reports of Tobacco Use? An Interactive Voice Response Experiment


Mayo, N. D., Currivan, D. B., & Muldavin, B. M. (2010, May). Does Voice Matter for Youth Reports of Tobacco Use? An Interactive Voice Response Experiment. Presented at AAPOR 2010, .


Household telephone surveys on youth tobacco use yield lower estimates than school-based self-administered surveys. Researchers generally assume the lower estimates from telephone surveys reflect underreporting due to youths’ concerns about parents or others overhearing their responses. Interactive voice response (IVR) has been shown to generally increase youth reports of smoking compared to interviewer-administered CATI surveys. Nevertheless, a significant gap remains between estimates from IVR and school-based surveys. One potential limitation of the IVR mode is that the “standard” human voice used to record the survey items is an adult female. Youth who respond via IVR may still feel like they are reporting to an adult, even though the voice is part of a computerized system. This may discourage some youths from fully reporting smoking behavior, as it could remind them that there is some risk, however small, of their smoking behavior being disclosed to their parents. For in-school surveys, the primary risk is that classmates who are also participating in the survey could see youths’ responses.

In order to assess what role, if any, concerns about disclosure play in youth smoking reports in IVR, we designed a survey experiment where youth respondents were randomly assigned to one of two different voices – an adult female who sounded old enough to be the respondents’ mother and a teen female who sounded close to the expected average age of respondents. To test whether the IVR voice had any impact on youth smoking reports, we compared responses across the two experimental conditions on two smoking behavior items. We also examined whether the IVR voice led to different response patterns for additional questions on youths’ intentions to smoke. This paper reports findings from the experiment and discusses the implications of the findings for measuring youth smoking in telephone surveys.