Mitigating Bias in Questionnaire Creation

Bias is a salient issue in the design of questionnaires, which contributes to inaccurate results in research. Often, bias arises because of an unanticipated communication barrier between respondents and investigators. The human factor plays a leading role in creating bias in the questionnaire, as the human mindset is often affected by how we perceive and respond to questions (Rowley, 2014). For instance, bias tends to occur due to language barriers, interpretational issues, and using complicated wording. During the questionnaire creation, the investigator will likely use ambiguous questions, uncommon words, jargon terms, or vague words that encourage vague and incorrect answers (Phillips et al., 2016). Also, the use of sensitive questions that ask about marital status, sexual orientation, household incomes, and age tends to elicit inaccurate answers. Bias can also occur when the questionnaire is not designed well; restrict the survey to people with telephones and not translated into all necessary languages (Nadler et al., 2015). Overall, bias in questionnaires occurs because of how investigators frame individual questions, administer the questionnaire, and design the questionnaire as a whole.

Collecting accurate data requires the investigator to design a questionnaire that is unbiased. Researchers are supposed to comprehend and be able to mitigate bias in the design of the questionnaire (Coombe & Davidson, 2015). One way to mitigate or at least minimize bias in the questionnaire is to carefully and precisely frame each individual question. Investigators need to avoid complex and lengthy questions and instead should focus on questions that are short, simple, and clear (Di Francesco, 2018). Complex and lengthy questions, such as double-barreled questions, increase ambiguity in the questionnaire and lead to response bias in the form of unfinished surveys or unclear answers (Booth, 2021). An example of a double-barreled question could be: Do you agree that COVID-19 can be transmitted by shaking hands with an infected person or through other means of physical contact? A “Yes” answer could mean yes by shaking hands or yes, through other physical contact means, or both. This type of question is likely to be answered wrongly, as compared to straightforward questions that are short and clear and are answered correctly and accurately.

Also, we can mitigate the issue of bias in the questionnaire by using precise, simple language. Bias in questionnaires is caused by the use of vague, uncommon, and difficult words and technical jargon (Aczel et al., 2015). An example of the use of jargon would be: “What was your age at menarche?” Instead of using simple words like, “What was your age when your menstrual period started?” Many women might not be familiar with the word ‘Menarche,’ thus they are likely to answer this question wrong. The use of obscure terms, company acronyms, and jargon needs to be avoided as they tend to confuse respondents, making them respond in a misleading or incorrect manner (Woolf & Edwards, 2021). Simple language yields precise answers.

Lastly, we can also mitigate questionnaire bias by avoiding leading questions. One cause of questionnaire bias is the use of leading or hypothetical questions (Lacroix et al., 2017). An example of a leading question would be, “Do you do physical exercise, like running?” This type of question will likely lead respondents to focus only on running. To mitigate bias, these types of questions need to be completely avoided.

Collecting accurate data requires the investigator to design an unbiased questionnaire. An unbiased questionnaire yields accurate and amazing results. Researchers are supposed to comprehend and be able to mitigate bias in the questionnaire's design. As discussed above, some ways to mitigate bias in a questionnaire include using precise, simple language, avoiding the use of leading questions, and using short, simple, and clear questions.

References

Aczel, B., Bago, B., Szollosi, A., Foldes, A., & Lukacs, B. (2015). Measuring individual differences in decision biases: methodological considerations. Frontiers in Psychology, 6, 1770.

Booth, C. (2021). Questionnaires. The Encyclopedia of Research Methods in Criminology and Criminal Justice, 1, 311-313.

Coombe, C., & Davidson, P. (2015). Constructing questionnaires. The Cambridge guide to research in language teaching and learning, 217-223.

Di Francesco, P., Lago, P., & Malavolta, I. (2018, April). Migrating towards microservice architectures: an industrial survey. In 2018 IEEE International Conference on Software Architecture (ICSA) (pp. 29-2909). IEEE.

Lacroix, E., Alberga, A., Russell-Mathew, S., McLaren, L., & von Ranson, K. (2017). Weight bias: A systematic review of characteristics and psychometric properties of self-report questionnaires. Obesity Facts, 10(3), 223-237.

Mackinnon, S. P., & Wang, M. (2020). Response-Order Effects for Self-report Questionnaires: Exploring the Role of Overclaiming Accuracy and Bias. Journal of Articles in Support of the Null Hypothesis, 16(2).

Nadler, J. T., Weston, R., & Voyles, E. C. (2015). Stuck in the middle: the use and interpretation of mid-points in items on questionnaires. The Journal of General Psychology, 142(2), 71-89.

Phillips, A. W., Reddy, S., & Durning, S. J. (2016). Improving response rates and evaluating nonresponse bias in surveys: AMEE Guide No. 102. Medical teacher, 38(3), 217-228.

Rowley, J. (2014). Designing and using research questionnaires. Management research review.

Woolf, B., & Edwards, P. (2021). Does advance contact with research participants increase response to questionnaires: an updated systematic review and meta-analysis. BMC Medical Research Methodology, 21(1), 1-27.