#jsm2018 Paul Schulz from @umisr up next, talking about mixed mode surveys on the Index of Consumer Sentiment from the Surveys of Consumers 
#JSM2018 Schulz Parallel modes for SoC - current mode is cell only; parallel study is ABS sample with mode choice for mail or web
#JSM2018 Schulz Web/mail is older, more female, lower income than phone. Can weight these differences away. But other differences may exist. Phone is considered benchmark in this study, so want to have web/mail match the phone.
#JSM2018 Schulz The technical details, but note the highly useful illustrative subscripts!
#JSM2018 Schulz Using sex by age by income as adjustment cells. Phone estimates for personal finances consistently above the mail estimates within each of these cells. other estimates have slightly different associations
#JSM2018 Schulz Predict cell-level mean for phone using age, sex, income covariates and web/mail cell mean. 3 approaches - OLS, WLS, and a measurement error model assuming both modes are random variables.
#JSM2018 Schulz The web/mail estimates are different from phone, but adjustments with the blended estimates seem to bring in line with phone and decrease SEs (larger sample size helps here)
• • •
Missing some Tweet in this thread? You can try to
force a refresh
