1
$\begingroup$

Can you correlate 2 Likert scales when the question for each one was different?

I have 1 question which asks how concerned someone is about pollution. They rank this on a Likert Scale of 1-5 where 1 is most concerned and 5 not at all.

Then my other question asks whether people intend to take any extra measures to reduce their own pollution in future, again on a scale of 1-5 but where 1 is very likely down to 5 not at all.

Can I run a correlation test on these two questions? Naturally I would expect the more concerned one is, the greater their intention to act. Obviously both scales are 1-5 but as the meaning behind what the number means is slightly different (I.e in Q1, 1= Most Concerned and in Q2, 1= Very Likely), can they still be correlated?

$\endgroup$
  • $\begingroup$ No problem, @s94, & you're welcome. If you register your account, you'll be able to come back & ask more questions when you need help, & you'll accrue reputation from your various questions. $\endgroup$ – gung Aug 20 '14 at 21:10
1
$\begingroup$

Yes, almost all correlations are run with variables that measure different (but hopefully related) things. You may prefer to use Spearman's correlation for Likert items.

$\endgroup$

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.