Survey Practice October 2010: Special Issue on Multi-Mode Surveys

This issue covers only one topic – multi-mode surveys. Diane O’Rourke selected the articles and edited the issue. The interest in this topic was evident when we received three times the number of submissions we could accommodate in this issue. We hope to include some others in future issues.

Using multiple modes for surveying has been common for decades – mail and other self-administered methods, face-to-face interviewing, then telephone interviewing. Now that Web-based methods are available to most potential respondents, there are many issues to consider.

The articles in this issue discuss differences by response mode – in respondent characteristics, question responses, cost and response rates, as well as differences by the timing of the modes. One article recommends a more detailed system for sample control in order to better understand the results of the multi-mode efforts.

Michael Brick and colleagues use the IRS Individual Taxpayer Burden Survey to look at differences in respondents by mode. They also explore nonresponse bias and ways to adjust for it.

Multiple modes can be used simultaneously or sequentially. Nicole Bensky, Michael Link and Chuck Shuttles present findings from multi-mode efforts to recruit people for Nielsen’s TV Audience Measurement Service.

Employee surveys can be administered in a variety of ways. Robert Teclaw and his colleagues analyze if Veterans Health Administration employees who select different modes differ on their demographic characteristics.

Christopher Friese and his co-authors present the results of an experiment that compares mail versus Web completion of a survey of nurses. This paper also explores the order of the modes as well as cost differentials.

Although face-to-face interviewing is expensive, Bart Buelens and Jan van den Braken of Statistics Netherlands discuss the implications of not including it in a mixed-mode survey in order to reach certain populations.

Sample control is an important yet often neglected aspect of survey administration. Now, more than ever, it’s important to track the results of multi-mode surveys by mode. David Chearo and Martha Van Haitsma present a set of attempt codes they have developed to track these results.

As always, please send your articles appropriate for Survey Practice and we welcome conference presentations.

The Editors

  • John Kennedy
  • Diane O’Rourke
  • David Moore
  • Andy Peytchev