Writing Survey Questions


Writing Survey Questions



This module discusses how to write survey questions and format responses.

Learning Objectives

  • Describe the different types of survey questions and give examples
  • Explain how to appropriately word and write good survey questions
  • Describe how to format the response options for questions
  • Discuss the importance of question placement within a survey

 

Writing good survey questions is a critical part of the survey process because the quality of the questions and the structure of survey will impact the reliability and validity of the data. There are two types of questions commonly used in surveys – open-ended questions and closed ended questions. Open-ended questions allow the respondent to respond to the question in their own words. Open-ended questions, also referred to as unstructured questions, allow for the greatest variety of responses and provide respondents with an opportunity to reflect on their answer. Sometimes unexpected answers may emerge that provide new insight in the research. The disadvantage is that responses from open-ended questions require more time and effort to analyze. Close-ended questions, also referred to structured questions, provide a list of acceptable responses and the respondents must choose the most appropriate response. The responses may include checklists, multiple choice, true/false, and attitude scales. These surveys are easier and less time-consuming for respondents and the data is easier to record and analyze for the researcher.

Responses to open-ended, unstructured questions are typically just short text responses. For close-ended, structured questions, however, responses may have a variety of specific formats. These may include:

  • Fill-in-the-blanks: These formats ask the respondent to check a blank and are often used for dichotomous responses where they select male/female, yes/no, and true/false. Another example would be when a respondent is asked to rank order or preference a list of choices by writing a number on every blank. There may also be questions that simply provide a blank and allow the respondent to fill in a one word response.
  • Multiple choice: Respondents will have a list of choices to choose from and will be asked to check mark or circle the most appropriate answer. This is called a single-option question because only one answer can be selected. A question listing multiple responses may also allow a respondent to check more than answer if more than one answer applies, making it a multiple-option format.
  • Attitude Scales: This format is like a multiple choice in that the respondent will circle a number or word that corresponds with their attitude or opinion about a particular topic. A scale that uses terms such as strongly disagree, disagree, neutral, agree, and strongly agree is commonly used. This is called a Likert Scale and there are several variations. Likert scales usually associate a number with the terms and use a 1-5 or 1-7 point rating scale.
  • Filter or contingency questions: In this type of questioning, the respondent’s answer to one question may determine whether or not they are qualified or experienced enough to answer a subsequent question. For example, a question may ask if the person is or has been a smoker. If the answer is yes, the respondent may asked a subsequent question regarding how many cigarettes per day he or she smoked on average.


Once the researcher has decided on the type and format of the question, he or she will have to write the questions for the survey. The structure and wording of the questions may impact the how the respondents answer the question. Therefore, careful consideration should be given to wording and how the question is written. The overall structure of the survey and the placement of the questions within the survey are also important. Following is a list of coomon tips for writing good survey questions and structuring the survey:

  • Surveys should begin with a clear set of instructions regarding how to answer the questions and a short “preamble” that sets the stage for the topics in the survey.
  • Start with general questions that are easy to answer and will be considered to be sensitive or threatening.
  • For historical demographics, it may be helpful to place the questions chronologically.
  • Ask about one topic at a time and avoid using words like “or” or “and” in questions. When switching topics, use a transition.
  • Make sure the question is not misunderstood and that the desired response type is obvious. For example if you ask what remedy they take to fight heartburn, do you want to know over-the-counter medications, prescriptions, home remedies, or any and all remedies?
  • Define terminology within the question if necessary to avoid assuming that the respondent will be able to effectively answer a question rather than “guessing” at a response.
  • Always specify a time frame, especially when using the words “will”, “should”, “may”, or “might” in a question.   For example, if you ask whether or not a person may be elected to a position, you would need to specify the time or date.
  • Avoid leading words/questions. For example, words like “force” have a strong meaning that may be construed negatively. If you asked whether or not parents should be “forced” to vaccinate their children, the word “forced” may influence how respondents answer.
  • Avoid wording that may seem biased or loaded. Sometimes this is avoided by asking about both sides of an issue.
  • Questions should be specific and direct. Example: If you asked customer at a restaurant how their experience could be improved, you would need to have separate questions to address wait time, food quality, service, etc.   If you were wanting know if someone liked a movie, it may be more telling to ask if they would recommend the movie to others, rather than simply asking if they liked the movie.
  • Avoid the use of jargon, technical terms, acronyms, abbreviations, symbols or another else that may make the question unclear.
  • Give mutually exclusive options for responses. For example, if asking age range, it would not be appropriate to use 0-10, 10-20, and so on. How would someone respond if they were 10? Instead it should say 0-9, 10-19, and so forth.
  • Be specific in response options. For example, if you want to know how often people go to church, do not use terms that are general and have different meanings to different people such as often and seldom. Ask for a specific number of occurrences within a time frame.
  • Make the layout of the survey visually appealing by using consistent fonts, colors and so forth. Also be consistent with the look of the question and the way the question responses are chosen.
  • Keep the questionnaire short to increase response rates.
  • Pay attention to the placement of questions in mind, starting with general questions and moving to more sensitive questions at the end. Be careful of not “priming” the respondents by mentioning something in an earlier question that may impact responses later.


The video below, Writing Good Survey Questions, provides additional information regarding types of survey questions and tips on how best to format the questions.



Suggested Readings

Aday, L. A., & Cornelius, L. J. (2011). Designing and conducting health surveys: a comprehensive guide. John Wiley & Sons.
Blair, J., Czaja, R. F., & Blair, E. A. (2013). Designing surveys: A guide to decisions and procedures. Sage Publications.
Converse, J. M., & Presser, S. (1986). Survey questions: Handcrafting the standardized questionnaire (Vol. 63). Sage.
Dillman, D. A. (2011). Mail and Internet surveys: The tailored design method--2007 Update with new Internet, visual, and mixed-mode guide. John Wiley & Sons.
Fowler Jr, F. J. (2013). Survey research methods. Sage publications.
Fowler, F. J. (1995). Improving survey questions: Design and evaluation (Vol. 38). Sage.
Thomas, S. J. (1999). Designing Surveys That Work! A Step-by-Step Guide. Corwin Press, Inc.
Woodward, C. A., & Chambers, L. W. (1983). Guide to questionnaire construction and question writing. Canadian Public Health Association.

Viewed 6,037 times