Questionnaires economically gather data from both large and small groups of people. Properly constructed, they do not take long to complete and statistical results can be quickly tabulated. Development of a questionnaire requires in depth planning, and usually more than one draft is necessary.
You may also have had occasions to respond to such questionnaires, sometimes in newspapers or sometimes from marketing personnel, who do door to door surveys.
Questionnaire design is critical. Questions should be short, easy to understand, unbiased, nonthreatening, and specific. To make sure questions will stimulate needed information, the analyst can test them with one or two outsiders before widespread distribution. Prepaid return envelopes accompanying questionnaires sent to outside help assure prompt response.
The analyst should send questionnaires to everyone involved with the system. A questionnaire works particularly well when the analyst must gather data from a large number of people, when the analyst must ask everyone the same questions, or when facts must be collected from people, such as suppliers, who do not work for the organization. Questions can follow four formats :
- Multiple choice : This gives respondents a specific set of potential answers. The format is ideal for computer tabulating.
- Open ended : Respondents must answer the question in their own words. Space is provided under each question for the response.
- Rating : This is similar to multiple choice except that respondents must rate their
- Rank : Rank requires respondents to prioritize their responses from high to low or on a percentage basis.
Aware that most people do not spend a lot of time responding to questionnaire. Most analysts decide to mix question formats, including follow-up questions, within the original questionnaire to permit elaboration of certain responses. By so organizing a questionnaire, the respondents have an opportunity to express their opinions freely, and yet answer
quickly through the use of multiple-choice, rating, and ranking questions. When all the questionnaires are returned, the data can be tabulated.
If the results of a questionnaire survey are incomplete or confusing, the analyst may want to contact selected outsiders by telephone or in person. This requires tact, of course, and an understanding that the analyst’s own pressing need may not concern outsiders in the least.