Perhaps a more qualitative approach may be useful for your purpose. Best wishes, David Booth

I'll assume that you can treat your variables as continuous variables. That is, that if each of Awareness and Attitude were measured with Likert items, that each is composed of several items and summed up. If this is not the case, some of the following may not be the best approach.

You want to use a paired test. That is, match the Awareness score of Person1 with the Attitude score of Person1. If you didn't record the information in a way to do this, then you might use a different approach.

Common paired 2-sample tests include paired t-test, Wilcoxon signed-rank test, and two-sample sign test. Each of these tests a different hypothesis, any one of which may be of interest to you.

I would also recommend plotting Awareness vs. Attitude. This may give you some insight into how the two are related.

For example, if you had a small data set:

Awareness = (1,2,3,4,5,6,7,8,9,10)

Attitude = (1,3,2,5,6,7,8,7,8,7)

You would not find a significant difference in e.g. the mean values, but the plot reveals an interesting relationship.

Hello Lawrence,

What you can do is, first, convert scores to a common scale. Z-scores will work just fine if you're using continuous scores (interval strength); ranks or percentile scores will work if ordinal scores or continuous scores are used.

Then, run your comparison. A within-subjects / paired measures approach would be best, as suggested by Salvatore.

Of course, all this hinges on the premise that it makes sense to compare persons on their relative standings on two different attributes.

Good luck with your work.