Does the respondent have the information Requested? People will answer questions they do not understand. Similar tests in which dummy advertisements are inserted in magazines, find people who report having seen the ads before. Such answers do not necessarily stem from dishonest. Many result from confusion. Results of this type emphasize the importance of asking only questions for which the respondents have the necessary information to answer. Several aspects of this point, which researchers should consider in planning their questions, are discussed in the following paragraphs.
Is the point within the respondent’s experience? Even questions that seem quite ordinary may not fall within a given respondent’s experience. How do you think ‘Tide’ compares with other packaged soaps and detergents? This question seems straight forward. Unless respondents have used Tide however, their answers are not apt to mean much. It would be better to determine in advance whether or not the respondent has ever used Tide. Even then some respondents will be comparing Tide with All and Era, while others will be comparing it with Ivory and Lux. It is important, then, to determine whether the question is meaningful in the respondent’s experience. When there is doubt ‘filter’ questions should be used. In this example, the researcher might ask: What packaged soaps and detergents have you used during the last year? If Tide is reported among others, the next questions might be: How do you rank these in order of your preference? The filter question sown here is indirect. It is probably preferable to a direct question in this case, as the latter might introduced some bias. If an indirect question is not feasible however a direct filter question should be used. For example: Have you used Tide during the last year? More people will say ‘no’ to such a questions than will volunteer that they have not used Tide if asked a question which assumes they have used it, such as, How do you think Tide compares with other detergents.
Can the Respondents remember the Information? How many can name the brand of shirt they are wearing what they had for dinner a week ago, or the amount of the deductions from their paychecks? Most people have known these at one time or another, but they are not apt to remember them when questioned specifically. Many things that one might expect everyone to know are not remembered. In fact, where memory is involved at all, researchers must be cautions not to overestimate the accuracy with which respondents will remember the information they want.
Consumers are particularly poor at remembering quantities of products used. Where factual data have been available for comparison, researchers have found consumer reports of product usage to exceed actual usage by 100 percent or more. Classification of consumers into heavy users and light users of a product on the basis of reported use has been found to misclassify as many as 77 percent of respondents asked about their consumption of tea and 41 percent asked about floor polish.
A great many questions asked in marketing research involve memory in varying degrees. What brand of soap to you use? Have you seen this advertisement before? What stories did you read in this magazine? What television programs did you watch yesterday? Memory of such events is influenced by four factors: (1) the importance of the event itself, (2) the individual being asked to remember, (3) the length of time since the event, and (4) the stimulus given to the individual’s memory. The first two of these factors are beyond the researchers’ control but they must be considered in deciding which questions to include in the survey. The last two factors researchers can effect.