Hello Zeeshan,

The choice of a response scale should likely be linked to: (a) the specific stimulus, question, or prompt presented to the respondent; (b) the intended population of respondents; (c) the intended type(s) of data analysis to be applied to the scores; and (d) whether you believe that a middle, or "neutral" point makes sense when measuring perceptions.

That said, it's probably fair to say that, for estimates of degree of agreement, endorsement, or the like (common in the measurement of attitudes, which was the domain in which Likert was proposing his scale method) that five options is the most commonly used.

There are those who assert that more option levels offers better score reliability and/or better conformance of rating scale scores to the behavior of a continuous variable.

See these links for more elaboration:

1. Kusmaryono, I., Wijayanti, D., & Risqi, H. (2022). Number of response options, reliability, validity, and potential bias in the use of the Likert scale in education and social science research: A literature review

International Journal of Educational Methodology, 8(4), 625-637.

2. Hutchinson, D., & Chyung, S. Y. (2023). Evidence-based survey design: Adding “Moderately” or “Somewhat” to Likert scale options Agree and Disagree to get interval-like data. Performance Improvement Journal 62, 17-24.

3. Simms, L. J., Zelazny, K. A., Williams, T., & Bernstein, L. (2019). Does the number of response options matter? Psychometric perspectives ssing personality questionnaire data. Psychological Assessment 31(4), xx-yy. doi:10.1037/pas0000648

Good luck with your work.

Zeeshan Baig, David Morse has given you some really useful information. Permit me to offer a slightly different perspective - and I hope it's OK that I move away from your original question.

Although many (most?) researchers use the Strongly disagree to Strongly agree continuum for response options, usually with something such as Uncertain in the middle, there is some evidence that these kinds of bipolar sets of response options are less satisfactory than unipolar sets, e.g., Never through to Always.

In research I have been involved in, there does seem to be benefit in using the second type of options.

The following articles might be of interest:

Chyung, S. Y., Roberts, K., Swanson, I, & Hankinson, A. (2017). Evidence-based survey design: The use of a midpoint on the Likert scale. Performance Improvement, 56(10), 15–23. https://doi.org/10.1002/pfi.21727

Nadler, J. J., 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. https://doi.org/10.1080/00221309.2014.994590

Saris, W. E., Revilla, M., Krosnick, J. A., & Shaeffer, E. M. (2010). Comparing questions with agree/disagree response options to questions with item-specific response options. Survey Research Methods, 4(1), 61–79. https://doi.org/10.18148/srm/2010.v4i1.2682

All the best for your research.

As Robert Trevethan's answer shows, there has been quite a lot of study of Likert-scoring by survey researchers. In general, the conclusion is that 5-point or 7-point response formats provide better results (e.g., higher reliability) than 3-point scores.

I don't think any further explanation is required after the above responses.

Just be mindful of the points David Morse mentioned:

"The choice of a response scale should likely be linked to: (a) the specific stimulus, question, or prompt presented to the respondent; (b) the intended population of respondents; (c) the intended type(s) of data analysis to be applied to the scores; and (d) whether you believe that a middle, or "neutral" point makes sense when measuring perceptions."

Robert Trevethan pointed:

"Although many (most?) researchers use the Strongly disagree to Strongly agree continuum for response options, usually with something such as Uncertain in the middle, there is some evidence that these kinds of bipolar sets of response options are less satisfactory than unipolar sets, e.g., Never through to Always."

And, David Morgan clarified: "In general, the conclusion is that 5-point or 7-point response formats provide better results (e.g., higher reliability) than 3-point scores."

The 5-point and 7-point Likert scales are widely used and balance granularity and simplicity well. The choice depends on your specific research objectives and the preferences of your target population. The following article, which could be of interest, explored and explained the Likert scale.

Joshi, A., Kale, S., Chandel, S., & Pal, D. (2015). Likert scale: Explored and explained. British Journal of Applied Science & Technology, 7(4), 396–403. https://doi.org/10.9734/BJAST/2015/14975

Good luck,

In addition to the very thorough answers above, I wonder if anyone has a comment as to the use of uneven vs. even numbered response scales.

I fear that the mid point in the uneven numbered scales may be ambiguous as it for some subjets may mean 'I am not sure what to answer, so I take the least informative' - sort of equal to 'I abstain answering this question' - while others may interpret it (as intended) as the mid point of the subjects evaluation of the issue.

Any comments?

Best,

Jan Ivanouw

Jan Ivanouw, the issue you raise has been considered helpfully in the following article that I included in an answer up above:

Nadler, J. J., 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. https://doi.org/10.1080/00221309.2014.994590

As I also mentioned up above, this problem doesn't arise in unipolar sets of response options such as Never through to Always.

Others might not agree, but I have found that participants also appreciate options such as Not applicable and Don't know.

Hello Dr Zeeshan Baig

I would intuitively think that having a wider range of options (eg 7-point scale) will lead to increased sensitivity. Would you (or anybody else) agree or disagree?

Joseph C Lee, your proposition seems logical, but there is evidence (forgive me for not looking it up to cite it in detail) that suggests having more than seven options does NOT increase "sensitivity". In some research I've been involved in, there can even be less sensitivity when there are more than seven response options - though some of that depends on what the options are.

The following might be of interest:

Trevethan, R., & Ma, K. (2021). Influence of response-option combinations when measuring sense of efficacy for teaching: Trivial, or substantial and substantive? Frontiers in Education, 6, Article 723141. https://www.frontiersin.org/articles/10.3389/feduc.2021.723141/full

Hello Zeeshan Baig, I think this really depends on the question and what validated and previous Likert scales have been used for the topics. I see your point though with seven categories being cumbersome, while three is easier but may not capture the true extent of the span of replies.