0
$\begingroup$

So I am trying to analyze my data. I used the following scales: Brief RCOPE, the Brief Multidimensional Measure of Religiousness/Spirituality (BMMRS), and the Satisfaction with Life scale (SWLS). They are all Likert type scales (1=dissatisfied, 2= slightly dissatisfied, etc.). Each measure has subscales, but I'm not sure whether I need to use the total scores for each subscale or the total score overall for each scale when I do correlations.

In SPSS, which statistical test should I use in order to find correlations between the Brief RCOPE + SWLS, and the BMMRS + SWLS (I Don't need a correlation between the Brief RCOPE and BMMRS)?

Also, should I be just summing up all of the scores for each measure or should I be using subscale scores? Because when I sum up the scores each person got on each measure, the data is all non-monotonic and I think that means I can't do some correlation tests like the Spearman's rank order. Thank you so much for any help you can provide!

additional info: all three variables are ordinal

$\endgroup$
0
$\begingroup$

Firstly, adding up and averaging the scores per measure per respondent does not 'exclude you from using spearmank rank correlation test', it ENABLES you to use a real correlation test like Pearson's. This is better (more power).

I think you should average scores for each measure (scale, set them to scale in SPSS), then test for correlation (by using Pearson's) 3 seperate times (AtoB, AtoC, BtoC)

(subscales I wouldn't use seperately unless they have a seperate function in your hypotheses) Hope this helped!

$\endgroup$

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.