Yes, you can change scale points. Generally it's recommended to keep similar scales, e.g. Likert-types, on the same point range.

A 6 point Likert scale sounds unusual, wouldn't it be unbalanced? A 5 point scale is most likely better suited for your research purposes.

Yes, you can change the scale points from a question to another in the questionnaire. But this is not a good idea, those who will answer your questionnaire will be more likely to get confused and maybe mix the instructions of different questions. This will result unreliable answers which may cause issues in the analysis later. Unless you have VERY strong reason for changing scale of the questions, you better go with a systematic questions. It is also advisable to unify the direction of the Likert-scale through the questionnaire: 1: Lowest ---- 5: the highest or vice versa, that is also a trick to pay attention to.

Dear Mrunali Damania,

I agree with the colleagues. Another point to think about is related no normality. When you have normality issues, it is better to have more points on your scale. Let me give you an example. Performance self-evaluation is a typical scale that people tend to answer at the end of the scale (i.e., of course I'm good, and I try my best!). In this case, when you have more points on your scale (7 or 9), you can catch a little bit more nuances than with fewer points (4 or 5).

Thus, although the original scale could be 5 points, you could have a reason to change/adapt your scale culturally as well. Some countries use an A, B, C, D, and E metric, another 0 to 20, others 0 to 10 or 1 to 10 or... etc. So, it depends a lot on where you are going to apply the measure and if the scale makes sense for them.

In this sense, making a semantic test of the scale and its range is interesting to see if you are not going to have problems afterward.

I'm going to suggest a chapter that, for me is brilliant that was written by Schwarz who is a very good methodologist. I know that it is in the consumer behaviour field that could be not yours, but the chapter is generalist. He is brilliant (well, I don't know him, I've just read his things, but they are always outstanding):

Weaver, K., & Schwarz, N. (2008). Self-reports in consumer research. In C. P. Haugtvedt, P. M. Herr, & F. R. Kardes (Eds.), Handbook of Consumer Psychology (pp. 1081–1102). Lawrence Erlbaum Associates.

I hope I could help you.

Scale points can be changed as per the nature of variable, educational background and experience of the respondents, and the maturity (Age) of the respondents. Scale points determine the sensitivity of the scale and must be decided based on the education qualification and experience of the respondents. uneducated or low educated respondents may not be able to differentiate between scale points if too many scale points are used. Same way children will also not be able to pick up finer variations in a variable, necessitating use of fewer scale points. The respondents must be able to differentiate between scale points, i.e. if the scale points are 1 to 5, the respondents should be able to respond accurately the variable value and they should not be confused in selecting one of the two consecutive scale points.

Thank you, everyone, for taking the time to answer my query.