U-M Surveys of Consumers adjusts its methodology, sentiment measures remain consistent

Over the next four months, the University of Michigan’s Surveys of Consumers will transition from its current method of interviewing American households via random digit dialing of cell phones to interviewing through web surveys with address-based sampling.

Web interviews will be gradually phased into the official data starting this month. By July 2024, all’interviews will be conducted over the web.

"Our research shows that estimates of key survey indicators are comparable across the two data collection methods,- said Joanne Hsu , director of the surveys. "The design of the methodological transition ensures the continued integrity of the time series data. It is informed by more than seven years of running a parallel data collection via the web alongside the official phone data collection, which demonstrates the continuity of consumer sentiment measures under both methodologies.-

Key survey features remain unchanged

Web-only surveys will maintain the key features of the current surveys. The data collection and twice-monthly release schedule will continue without any changes. Data dissemination methods remain the same; data users will be able to download and access survey data in exactly the same manner and schedule as they are currently.

All core interview questions, including qualitative items, have been adapted for the web and designed for both personal computers and mobile devices. The survey will remain a rotating panel. The target population remains adults aged 18 and older residing in the coterminous United States.

Current transition continues legacy of evolution based on hands-on research and experimentation

Over the past seven decades of the Surveys of Consumers, various challenges with data collection-and technological improvements in survey methodology-have arisen periodically. The surveys have addressed these challenges through regular improvements in methodology.

For the first 25 years, interviews were conducted face to face. Landline telephone interviews began in 1978. In 1993, interviewers switched from using paper questionnaires to using computers. Cell phone interviewing was phased in starting in 2012, and landline interviewing was phased out by 2015. As such, the transition to web interviewing continues the surveys’ long legacy of technological evolution.

The adjustments to the survey methodology are supported by 14 years of in-house experimental research with web interviewing and seven years of parallel collection of web data by the Surveys of Consumers. The methodological design of the transition is consistent with the recommendations of the American Association for Public Opinion Research Task Force on Transitions from Telephone Surveys to Self-Administered and Mixed-Mode Surveys.

Technical details about the transition are posted on the Survey Information tab of the Surveys of Consumers data website.