Demographic and Measurement Differences between Text-to-Web and Phone Survey Respondents

This paper builds on existing literature on survey mode effects. We explore the relationship between administration mode and demographics as well as measurement issues. Participants in a low-income home weatherization program were assigned to either a phone call or text-to-web survey administration group. Our findings appear to be consistent with previous research regarding both non-observation and observation effects to varying extents. In terms of non-observation mode effects, we found that text-to-web and phone group survey respondents had similar demographic and home characteristics. The two survey methods yielded similar response rates and minimal statistically significant differences between respondents’ reported background characteristics. We found a larger portion of phone respondents chose “Prefer not to say” for some demographic questions and generally indicated higher satisfaction than text-to-web respondents.