0

I have two different data sets, one of them is a 7-level Likert scale and another is a 5-level Likert scale. In addition to the different number of levels, the 7-level Likert scale has 114 items and 13 subscales, and the 5-level Likert scale has 64 questions and 7 subscales.

Is it possible to calculate correlation among these mismatched subscales? The inventories are measuring thinking styles and learning strategies.

jonsca
  • 1,772
SaR
  • 1
  • I am using 3 independent variables( Recent Terrorists attacks,Past trauma, Major Life stressors) 3 dependent variables ( PTSD, Depression & Anxiety)and 2 moderating variables(Self Efficacy and Social Support). All variables have different Likert Point Scales. If I take average of each variable separately and then run regression analysis, is it Ok. What will be accuracy level of this regression analysis? – tariq Jan 18 '14 at 16:31
  • Welcome to the site, @tariq. This isn't an answer to the OP's question. Please only use the "Your Answer" field to provide answers. If you have your own question, click the [ASK QUESTION] at the top of the page & ask it there, then we can help you properly. Since you are new here, you may want to read our tour page, which contains information for new users. – gung - Reinstate Monica Jan 18 '14 at 16:33

1 Answers1

2

If the subscales are sums of several (or more) Likert items, then you can calculate correlation in the ordinary way, using Spearman's or Pearson's as appropriate, depending on the pattern of the scales.

If you want correlations among the Likert scales themselves, or if a subscale is the sum of only a couple of Likert scales, then polychoric correlations may be more appropriate; these assume that the ordinal variables are representations of an underlying latent trait that is continuous.

Peter Flom
  • 119,535
  • 36
  • 175
  • 383