0
$\begingroup$

I am looking for a statistical method which measures the relationship between two variables which are based on different likert scales and are independent from each other.

The first variable is a score which represents the pro-environmental behavior (PEB) (Likert Scale A)

The second variable represents the pereceived persuasiveness towards selected game design elements (Likert Scale B)

My aim is to show a relationship between the pro-environmental behavior score and the perceived persuasiveness (e.g. - Do people with a higher PEB - score perceived certain game design elements differently?)

I was thinking about a simple pearson correlation?

Furthermore, I have age and education, which may influence the PEB - score as well (this I want to show in a second step - probably through a multiple regression analysis).

Thank you.

$\endgroup$
1
  • $\begingroup$ agree with spearman rho. other thing you could look at is ordinal regression (for your multiple regression analysis) $\endgroup$
    – seanv507
    Commented Mar 29, 2019 at 13:26

1 Answer 1

0
$\begingroup$

As you are using likert scales which are ordinal rather than continuous I'd suggest using a spearmans rho as it can be used on ordinal data unlike pearson correlation

$\endgroup$
7
  • $\begingroup$ Also if I took a mean based of the likert scales? $\endgroup$
    – mboeckle
    Commented Mar 29, 2019 at 10:41
  • $\begingroup$ Yes you shouldn't really be treating a Likert scale like a continuous one. A lot of researchers do. But it's not correct. It'd be like trying to get the mean of 1st, 2nd, 3rd, 4th place. It wouldn't make sense. $\endgroup$
    – A Soutter
    Commented Mar 29, 2019 at 11:09
  • $\begingroup$ Anything else I could possibly use appart from this? $\endgroup$
    – mboeckle
    Commented Mar 29, 2019 at 11:18
  • $\begingroup$ I mean a spearmans rho is as easy to do as a pearson's one and is interpreted in much the same way. $\endgroup$
    – A Soutter
    Commented Mar 29, 2019 at 11:33
  • $\begingroup$ Thank you, do I need to normalize it? $\endgroup$
    – mboeckle
    Commented Mar 29, 2019 at 11:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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