A regular client asked me to do a couple of pretests in the last few months. Pretesting used to be much more common among polling organizations than it is now. The idea is to do about a dozen interviews using the final draft of the questionnaire. We time each interview to calculate the average length. The client expects me to report any errors in language or logic of the questionnaire. They also expect me to report any problems with the sample.
I feel fortunate that I have a client who thinks of not pretesting as a penny-wise and pound-foolish proposition. Even if he does not need to rewrite questions, he finds pretest data helpful for accurate proposals and budgets. We always learn something from a pretest.
One thing I like about getting on the phone to do a pretest is that we often learn about hot-button issues we didn’t know about. Surveying people about a school levy, park levy or library levy sounds dry, but questions on a questionnaire sometimes provoke excited responses on educational policy or environmental issues. When this happens, a pretest provides a good excuse to let a respondent go off topic. Doing so can illuminate issues that have not been addressed by either side of a levy issue.
Pretesting also provides an opportunity for me to evaluate my interviewing skills. Respondents let me know pretty quickly if I read questions too quickly or too slowly. The better I get at listening, the earlier I learn which probes yield useful information for the client.
I would be glad to have a chance to help other opinion research or market research organizations with pretests, if there are any that still do. If so, please let me know.
John C. Stevens
(614) 772-2332
