I have a question on coding a Likert scale in SPSS.

Assuming we have a 7-point Likert scale, a "well-being" score may rank from 1 through 7.

While doing some reading on multi-variate analysis, I've come across a reference that suggests a "deviation score". That is, it's ranging from [-3, -2, -1, 0, 1, 2, 3].

For later analysis of survey results, which method ("well-being score" vs. "deviation score") will serve me best? Or are these "apples and oranges"?

Thanks, EEH


More like apples and apples.

For most purposes, it won't make any difference in the results, only in how they are interpreted. That can be important. One exception is if you wanted to add a quadratic of the Likert scale.

If you tell us more about what you are doing (e.g. is this the dependent variable? Independent variable? Or just something you are describing (e.g. with frequencies, median, etc) we will be able to give more detailed advice.

  • $\begingroup$ All "extent" question are referenced to the independent variables. There will be a few question for the dependent variables; however, I won't apply Likert scales for the DV-related questions. Based on the answers from the IV questions, I'd like to proof/disproof the hypotheses. Did that help? $\endgroup$
    – user22000
    Mar 18 '13 at 14:19
  • $\begingroup$ Sure, that makes me more sure that my answer applies to your case. $\endgroup$
    – Peter Flom
    Mar 18 '13 at 14:27
  • $\begingroup$ Thanks... could you please elaborate on "... only in how they are interpreted. That can be important."? $\endgroup$
    – user22000
    Mar 18 '13 at 16:01
  • 1
    $\begingroup$ It's more important when it's the DV, I think. When it's the IV in a regression, you are mostly interested in how a change affects the DV. $\endgroup$
    – Peter Flom
    Mar 18 '13 at 17:30

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.