I used a -4 to +4 Likert scale in one set of questionnaires and -3 to +3 for the same questionnaires, but gathered respondents in a different way. I want to rescale from -4 to +4 down to -3 to +3 for the second group.

How do I do this and secondly is there a referenced/accepted procedure for doing this?

  • $\begingroup$ I am also curious, in rescaling is it easier to change '1' (slightly agree) into a '0' due primarily to '1' measuring a smaller effect than from say a '4' (totally agree) to a '3'? $\endgroup$ – hucsy Oct 1 '11 at 12:46
  • 2
    $\begingroup$ I believe similar questions have been asked in the past, see How to compare Likert scales with varying number of categories across time?. $\endgroup$ – Andy W Oct 1 '11 at 13:24
  • $\begingroup$ Hmm okay. Another question then. The questionnaire I used has literature to say that depending on the sample populations age and capacity, the scale can be altered to -2 too +2. This leads me to wonder then, given I am only really interested in a final value for each person (after calculating SS Stage/ SSDev x 100), do I actually need to make adjustments? Why would the author say that if it was going to impact the final value! Its been 10 years since I did stats (hence this issue) so I am struggling to find an understandable, known option. $\endgroup$ – user6613 Oct 2 '11 at 13:44
  • 1
    $\begingroup$ Hi hucsy, if you register your account you should be able to edit your original question. $\endgroup$ – Andy W Oct 3 '11 at 12:19

Rescaling the variable as if it were continuous is easy enough. If it's centered on zero and the end points represent real extremes you can just just scale something like:

$$X_{SCALED} = X_{ORIG} \times MAX_{SCALED}/MAX_{ORIG}$$

That said, as was suggested in the answer to the other question linked in the comment above, scaling Likert scales is, for substantive reasons, not a great idea. Simple scaling with you result in non-integer answers which are impossible on either scale and might distort the substantive meaning that comes from scales where numbers represent real statements (e.g., 2 means "somewhat satisfied" while 0 means "neutral").

I'm generally hesitant to treat Likert scales as continuous in the first place. Look carefully at the distribution of answers on your two different scales to get a sense of any distortion that you have. It may make more substantive sense to do something other than scaling (i.e., collapsing categories in one or more of the scales and then comparing the variables as categorical).

  • $\begingroup$ I guess the point is also if we're talking "Likert scale" or "Likert item" here. Scaling individual items is super-bad since they already have enough problems with validity as it is. I guess rescaling a Likert scale (i.e. an aggregate of multiple Likert items) may be okay since there are also enough people who treat Likert scales as interval data. $\endgroup$ – xmjx Oct 1 '11 at 16:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.