Making key marketing decisions based on results obtained from inadequately designed research questionnaires could expose you to the possibility of significant profit loss. The phrase, “garbage in, garbage out,” applies here, meaning the questionnaire provided nonusable or misleading results. A well-designed questionnaire will provide valid data that will lead to actionable findings.

All surveys require a questionnaire that asks the right questions in the right order in the right way to the right people. Here are some rules to keep in mind when designing a questionnaire.

• Do a little homework. Take all the information on the topic to be surveyed from clients, do some investigating of your own (publications, sales people, other available data) and design a preliminary questionnaire.

• Never use the word “and” in a question or answer. Why? Because you are asking and respondents will be answering two questions. Now you have problems in interpretation and tabulation.

• Ask the necessary qualifying questions to be certain the person being interviewed is qualified to participate in the survey. Answers to questions regarding product usage, age and gender can help you be certain the data being generated comes from a qualified person.

• Make answer scales appropriate for the question. In answer categories, make sure the choices have distinct meaning. If respondents say a company provides good service, are they referring to customer service or product servicing?

• Use language that is appropriate for the audience. Use terms your audience understands and uses.

The only way to develop a sound questionnaire that provides solid information is by pretesting until the flaws and errors are minimized or eliminated. The primary reason questionnaires are “the weakest link” most often results from inadequate pretesting and not understanding that only pretesting will expose the flaws and errors, so they can be corrected.

By administering the questionnaire to a few people, pretesting can identify the problem areas. Always pretest with people who would qualify to be included in the survey.

After the pretest, talk to each person interviewed and note his or her overall responsiveness, comfort with the wording and ease of responding. Revise the initial questionnaire accordingly and retest.

Here are some thoughts to keep in mind when pretesting:

• Are the questions understandable? Did people ask what was meant by a question? If that happens, then the question should be rewritten to be certain it is understood.

• Does each question logically follow the last question?

• Is a question impossible to answer? Does everyone answer, “don’t know” in the pretest?

• Were any questions misleading?

• Were there any questions that required specific answers– that the respondent might or might not have? If the answers are not readily obtainable, the respondent might give up and abandon completing the interview.

In summary, remember to define specific objectives and make sure you are able to answer the basic research question with the responses obtained from your questionnaire results.

Select the right market-research methodology (for instance, mail, telephone, Internet, etc.) to administer your survey and pretest the questionnaire. Otherwise, you may interview 25, 250 or 2,500respondents only to find out you asked a key question incorrectly so the answers you get have little or no meaning. Don’t let the survey become the weakest link so that your findings are less than valid.

Earl Wilson is president of Wilson & Associates, a full-service marketing research company. Wilson & Associates offers Sales Lead Report™ subscribers a special 15 percent discount for initial research projects. For more information, email wilsonresearch@msn.com  or call 818-340-6201.