Hints for Designing Effective Questionnaires by Robert B

“Hints for Designing Effective Questionnaires” by Robert B. Frary Presentation by Brandon Benitez

Introduction Be Brief and Concise Ask for Feedback Put Personal Questions at the end Use categories properly Use Positive and Negative Stimuli in questions Do not use ambiguous inquiries Avoid category or scale point proliferation Avoid inquiries requiring ranked responses

Be Brief and Concise Use only questions that are necessary to collect the desired data. Having too many questions could annoy potential responders and lead to answers that are not completely thought out.

Ask for Feedback Ask for feedback from a representative sample of the responders. Consider using a preliminary questionnaire in order to test the effectiveness of it.

Place any personal questions at the end Locating personal and confidential questions in the middle or beginning could make potential responders uncomfortable. Locating such questions in the wrong places could also cause responders to give incomplete answers or no answers at all if they are unwilling to respond to them.

Use Categories Properly Be sure to place lowest order categories on the left and highest on the right if the categories are scalable. An example of this would be using “ 1)Never or Seldom 2)Sometimes 3)Always”. Consider condensing a large number of categories into fewer, more general ones if possible for the responder’s convenience. Use relevant categories and appropriate language to describe them. For example, use the phrase “Tend to Agree” in a scaling point inquiry so responders are allowed a degree of uncertainty in their answers.

Use Positive and Negative Stimuli in Questions Using only positive or negative inquiries alone can responders to give more unilateral answers before considering all questions. Having a healthy balance of both stimuli encourages the responder to be more thorough in choosing an answer.

Do not Use Ambiguous Inquiries Avoid the use of inquiries with open ended answers because not every responder is willing or able to provide them. Be careful not to provide the response option “other” unless it is essential since responders may choose that option carelessly. Avoid the inappropriate use of neutral response positions such as “undecided”, because it is very difficult to know the reasons why a responder chose that position. Design flexible options which responders are not required to make a strong stand, such as “tend to agree”, to help alleviate the problem.

Avoid Category and Scale Proliferation Design inquiries with as few categories as possible for answers unless there is a specific reason to have more. Providing too many categories on an inquiry is unnecessary and the results are oftentimes not useful. Avoid inquiries with a lot of scaled responses to choose from. An example would be rating something on a scale of one to ten. Using too many scaling points can often lead to annoyance, and the article mentions that responders can only be expected to decide between four or five choices at a time.

Avoid inquiries requiring ranked responses. Writing inquiries that require a user to rank various items, such as choice of paint colors, can easily lead to misinterpretations if the directions are not clear. Using such inquiries can also lead to annoyance as well. Responders can only be expected to rank up to six items at a time according to the research conducted in this article.

Conclusion The article provides these pointers in order to point out common pitfalls in designing questionnaires. It also gives some suggestions for alleviating these problems and to receive more reliable answers from responders. It does note that sometimes the use of certain types of inquiries is required regardless of these suggestions. It all depends on the purpose of the questionnaire itself.

Questions?