Hi Zeina!

First, it would be interesting to know how you measured both scales. Did you use Likert-Scales? If you have a high range of values, you could do a median split (that's working for not normally distributed datasets as well) => Split the sample into two samples (above and below the median, with one of the samples including the median, of course). Is this of any help?

Cheers,

Andy

I am looking for the answers. Because it is a good question.

Good luck.

I am not a specialist in statistics, but I think it can calculate it by finding the difference between the highest and lowest score and then dividing the result 3 where the division result represents the range for each class.

Check this article might be helpful for you.

Article Emotional intelligence as a predictor of self-efficacy among...

Dear Zeina

There are several methods of doing this. Assuming that you have used a Likert scale. The ones below P25, ones between P25 and P75 and the other group above P75. That is ones below the first quartile could be one group. The middle group would be the ones whose scores are between first and third quartile while another extreme group would be the ones above the third quartile.

You can see how we did with these variables:

You can use cluster analysis. You can check the attached article for the method