News Releases

Contact: Jon Meerdink ([email protected])

ANN ARBOR — Finding ways to increase survey participation is a crucial concern for social scientists whose work depends on respondents’ willingness to answer survey questions. In a new paper from a team at the University of Michigan’s Institute for Social Research, researchers have identified several ways to increase survey participation, giving survey designers new tools for gathering important data.

“Methods for improving participation rates in national self-administered web/mail surveys: Evidence from the United States,” published in August via PLOS ONE, used a nationally representative self-administered survey to test several hypotheses related to survey participation. Collecting information on many of the same variables as the National Survey on Family Growth, the new survey, which used both web and mail data collection, explored different styles of survey design and follow up, some of which resulted in increases in participation rates. Researcher Brady West said the self-administered nature of the survey was ideal for testing these hypotheses. 

“One of the drawbacks of surveys that use the web and mail relative to face-to-face interviewing is that they tend to get smaller response rates, and that introduces a risk of selection bias in the estimates that you’re computing,” he said. “So what this paper is really looking at is whether we can recruit people, especially if you’re looking at a very specific target population, like we had ages 18 to 49. Are there clever ways that you can recruit people and convince them to participate in a study?”

West and his co-authors on the paper utilized several incentives to increase participation. At the screening phase, they used a priority mail follow-up after an initial mailed invitation to participate via the web. The initial mailed invitation included a $2 bill as a gesture of goodwill, and a later priority mail follow-up included an additional $5 in hopes of convincing participants to complete an initial household screening. The combination of the priority mailing and the additional incentive was found to increase participation rates.

However, an even more effective method included telephone follow-ups for reluctant participants. When possible, the research team used commercial telephone data for households in the sample, and if a phone number was available, they attempted to contact initial nonrespondents to the web invitations by phone to simply remind these individuals about the study (and not actually do the interview over the phone). According to West, these telephone reminder efforts increased participation rates by about 20 percentage points.

“That was a big finding. I didn’t think it would be that effective to try to make contact with reluctant participants via commercially available phone numbers and simply remind them about the study if we made contact.”

The findings come at an opportune time. The COVID-19 pandemic has emphasized a need for survey methods that don’t depend on face-to-face contact with interviewers, which is already a costly and time consuming way of conducting research. In addition, certain survey methods like self-administration are better suited to collecting data on sensitive subjects, such as rates of forced sexual intercourse.

“I think people are slowly becoming a little bit more comfortable with participating in web surveys as opposed to participating in face-to-face or telephone surveys, provided that they’re reassured that their responses are going to be kept confidential,” said West. 

The full text of the study is available from PLOS ONE. Funding for the study was provided by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) of the National Institutes of Health.