The general format of the questionnaire is easy to overlook, yet it is something that is just as important as the wording of the questions asked. A questionnaire that is poorly formatted can lead respondents to miss questions, confuse respondents, or even cause them to throw the questionnaire away.
First, the questionnaire should be spread out and uncluttered. Oftentimes researchers fear that their questionnaire looks too long and therefore they try to fit too much onto each page. Instead, each question should be given it’s own line. Researchers should not try to fit more than one question on a line because that could cause the respondent to miss the second question or get confused.
Second, words should never be abbreviated in an attempt to save space or make a questionnaire shorter. Abbreviating words can be confusing to the respondent and not all abbreviations will be interpreted correctly. This could cause the respondent to answer the question a different way or skip it entirely.
Lastly, ample space should be left between questions on each page. Questions should not be too close together on the page or the respondent might be confused as to when one question ends and another begins. Leaving a double space between each question is ideal.
Formatting Individual Questions
In many questionnaires, respondents are expected to check one response from a series of responses. There may be a square or circle next to each response for the respondent to check or fill in, or the respondent might be instructed to circle their response. Whatever method is used, instructions should be made clear and displayed prominently next to the question. If a respondent indicates their response in a way that is not intended, this could hold up data entry or cause data to be miss-entered.
Response choices also need to be equally spaced. For example, if you’re response categories are "yes," "no," and "maybe," all three words should be equally spaced from each other on the page. You do not want "yes" and "no" to be right next to each other while “maybe” is three inches away. This could mislead respondents and cause them to choose a different answer than intended. It could also be confusing to the respondent.
The wording of questions and response options in a questionnaire is very important. Asking a question with the slightest difference in wording could result in a different answer or could cause the respondent to misinterpret the question.
Oftentimes researchers make the mistake of making questions unclear and ambiguous. Making each question clear and unambiguous seems like an obvious guideline for constructing a questionnaire, however it is commonly overlooked. Often researchers are so deeply involved in the topic being studied and have been studying it for so long that opinions and perspectives seem clear to them when they might not be to an outsider. Conversely, it might be a new topic and one that the researcher has only a superficial understanding of, so the question might not be specific enough. Questionnaire items (both the question and the response categories) should be so precise that the respondent knows exactly what the researcher is asking.
Researchers should be cautious about asking respondents for a single answer to a question that actually has multiple parts. This is called a double-barreled question. For example, let’s say you ask respondents whether they agree or disagree with this statement: The United States should abandon its space program and spend the money on health care reform. While many people might agree or disagree with this statement, many would not be able to provide an answer. Some might think the U.S. should abandon its space program, but spend the money elsewhere (not on health care reform). Others might want the U.S. to continue the space program, but also put more money into the health care reform. Therefore, if either of these respondents answered the question, they would be misleading the researcher.
As a general rule, whenever the word and appears in a question or response category, the researcher is likely asking a double-barreled question and measures should be taken to correct it and ask multiple questions instead.
Ordering Items In A Questionnaire
The order in which questions are asked can affect responses. First, the appearance of one question can affect the answers given to later questions. For instance, if there are several questions at the beginning of a survey that ask about the respondents’ views on terrorism in the United States and then following those questions is an open-ended question asking the respondent what they believe to be dangers to the United States, terrorism is likely to be cited more than it otherwise would be. It would be better to ask the open-ended question first before the topic of terrorism is "put" into the respondents’ head.
Efforts should be made to order the questions in the questionnaire so they do not affect subsequent questions. This can be hard and nearly impossible to do with each question, however the researcher can try to estimate what the various effects of different question orders would be and choose the ordering with the smallest effect.
Every questionnaire, no matter how it is administered, should contain very clear instructions as well as introductory comments when appropriate. Short instructions help the respondent make sense of the questionnaire and make the questionnaire seem less chaotic. They also help put the respondent in the proper frame of mind for answering the questions.
At the very beginning of the survey, basic instructions for completing it should be provided. The respondent should be told exactly what is wanted: that they are to indicate their answers to each question by placing a check mark or X in the box beside the appropriate answer or by writing their answer in the space provided when asked to do so.
If there is one section on the questionnaire with closed-ended questions and another section with open-ended questions, for example, instructions should be included at the beginning of each section. That is, leave instructions for the closed-ended questions just above those questions and leave the instructions for the open-ended questions just above those questions rather than writing them all at the beginning of the questionnaire.
Babbie, E. (2001). The Practice of Social Research: 9th Edition. Belmont, CA: Wadsworth/Thomson Learning.