My Survey Research Resolutions for 2010
Posted by Jeffrey Henning on Fri, Jan 01, 2010
Since I give advice to others about survey research, I'm always embarrassed when I don't follow my own recommendations. With that in mind, I've made five resolutions for my survey research in the New Year.
- Shorten my surveys! Long surveys become tedious and lead to bad results, since long surveys make liars of respondents. Worse, they make it less likely than respondents will take the next survey I invite them to. So I'll be following these six tips for shortening questionnaires.
- Test, test, test the surveys. Nothing is less professional than sending out a survey that doesn't work the way I intended - requiring answers to questions that aren't applicable, routing respondents through skip patterns that don't make sense, omitting common choices from choice lists. Self-test, pre-test, pilot test!
- Send out fewer reminders. Yes, survey reminders can double response rates, but many times a low response rate produces accurate enough results for the decisions at hand. I need to stop bothering people because of my past hangup about high response rates (recent research indicates that there is little correlation between low response rates and lower representativeness).
- Share the results of my surveys with respondents. People who spent the time answering my questions certainly deserve to have me close the feedback loop: sharing a summary of the results and letting respondents know what actions we are taking based on those results.
- Answer other people's surveys. While some surveys, especially access-panel surveys, screen out market researchers such as myself, I certainly owe the others the time to answer their questions. After all, I want people to answer my surveys, so I need to answer their surveys as well. It's a matter of survey karma!
Happy New Year! Good luck with all your resolutions, research-related and otherwise!