When designing a questionnaire, you need to know the right way to scale your items. I'll explain why using a five-point Likert scale is better than using a seven-point Likert scale, as well as how it impacts your results.

Five-point Likert scales are more bottom-heavy. The responses tend to be concentrated toward the lower end of the scale.

Five-point Likert scales are more bottom-heavy. The responses tend to be concentrated toward the lower end of the scale. For example, if you asked people how much they liked your product or service, they would give 1 or 5 (1 is not very much and 5 is extremely much). If you asked them if they would recommend it to others, they might select 4 or 5 (4 means that it's okay but nothing special - so no recommendation).

Seven-point Likert scales are more top-heavy. The responses tend to be concentrated toward the higher end of the scale.

Seven-point Likert scales are more top-heavy, as responses tend to be concentrated toward the higher end of the scale. They're used when there's a need to measure attitudes or opinions.

A five-point Likert scale is skewed to the left and a seven-point Likert scale is skewed to the right.

A five-point Likert scale is skewed to the left and a seven-point Likert scale is skewed to the right. This is because people tend to select at the extremes of the scale, so you should use a seven-point scale for both your questions and responses. The middle of a five point Likert scale (1–5) has been shown to be unhelpful when it comes to understanding how someone feels about something; instead, people tend to opt for extremes on both ends of each category (i.e., 1 or 5). For example:

• If you ask someone how their day went, they may answer “good” with their eyes closed while holding up one finger; this person probably did not have an especially good day but they could also say that everything was terrible!
• If you show someone two images and ask them which one looks better than another (with no other context), they will likely give an extreme answer like “this one looks more realistic” even though both images are actually quite similar in terms of color/lighting etcetera…

If you have a five-point Likert scale, then people will typically only select at the extremes of your items (1 or 5). If you have a seven-point Likert scale, then people will typically select in the middle (4 or 5) of your items.

So if you think about it this way, it makes sense that we’re more likely to be able to answer questions about ourselves when there are fewer options than when there are more options.

Conclusion

Likert scales are an excellent way to collect feedback from your customers, but they’re not perfect. This can be a problem if you want to get a wider range of responses than just two extremes. In that case, it might make sense to use a seven-point scale instead of five or seven points on your survey @ Nishadi Gampathirana

Hope this helps thanks!

Sunday Siman , any reference for this information ?

Nishadi Gampathirana , I'm no expert in this, but I suspect that a 10-year-old will be able to understand 7-point Likert items. If you are able to format the questions visually, I think the images like at the following would be helpful for adolescents: Article Diminished Social Motivation Negatively Impacts Reputation M...

Thank you Sunday Siman and Salvatore, for your valuable insights.

Again many many thanks Siman, for your comprehensive answer. If you have any recommended article, please share it to me.

Nishadi Gampathirana, for what it's worth, colleagues and I have used 7-point response options in research and have found that the distribution of responses on almost all of our items conformed pretty much to the normal curve, as did composite scores across all, or subsets of, items.

We used unipolar response options (e.g., never through to always), and that might have helped.

I have seen results from colleagues who used 5-point bipolar options that ranged from strongly disagree to strongly agree, with something like "neither agree nor disagree" in the middle. Their results have often concerned me, mainly because the middle option might create the impression of normally distributed data whereas it more likely is used for a variety of responses, including don't know, don't care, not applicable, a little bit of each, and a lot of each.

I think that it's best to have 7 response options, most or all of which are labelled, and two extra options of "Don't know" and "Not applicable".

All the best with your research.

7 scale

Thank you Robert and Edosa

Hello,

the choice of the number of response categories for a likert scale may depend on several variants of a search. I recommend that you read chapter 5 of Frikha Azza's book (2019), Measurement in marketing: operationalization of latent constructs, co-edition ISTE-Wiley, London UK-Hoboken USA.