When to Kick Out a Survey Respondent
Nearly every survey begins with screening questions to ensure that only the people you are trying to reach are included in the survey. For example, if you are conducting a survey of women, you need to ask about gender and kick out the men. And because every question costs money, you want to qualify respondents quickly and terminate those who do not belong.
Here is a helpful hint: Do not actually terminate respondents until after you have asked all screening questions. Suppose you’re trying to reach women over age 30 who own dogs. You’ll need three screening questions: gender, age, and dog ownership. First you ask gender. It’s tempting to have men immediately discontinue, and then after asking age have all those under 30 discontinue, and so on. Do not do this. Ask gender, age, and dog ownership of everyone. You don’t need men in your survey, but knowing their age and whether they own dogs may be useful to you. Likewise, you don’t need young women, but knowing whether they own dogs may be useful. We see two common scenarios in which this information becomes extremely useful:
1. You complete the study, present the findings, and somebody asks, “What if we expand our market to men?” Having complete screening data lets you estimate how many men there over age 30 who own dogs, which is an important piece of exploring whether to expand the market.
2. You decide to expand your target group while still in the field, which means relaxing the requirements of who qualifies for the survey. “Let’s open up our survey to allow women who are age 25 and over,” your client suddenly says. Having complete screening data tells you in advance how many more women will qualify for the survey and helps you estimate in advance the potential cost and benefit of relaxing this criterion.
Programming a survey to do this is easy, but unfortunately most people don’t. Give your programmer clear instructions. The additional cost of asking all screening questions to all respondents who want to take the survey is minimal, and the benefits are potentially huge. It’s a smarter way to screen respondents.
Need additional help? Versta Research would be happy to review your survey and assist in any way we can.
—Joe Hopper, Ph.D.