Correlation and regression analysis with survey answers
Hello dear community,
I have to execute a correlation analysis (Spearman) and linear regression analysis from a survey I did.
Now I have (for example) 20 answers to the topic job satifaction and 7 answers for the topic engagement.
Thus, in the table it says:
js1 js2 js3 ... e1 e2 e3...
When I do the analysis, it analyses every answer with every answer, however, I need to analyse the correlation of the "topic block" satisfaction with the other "topic block" engagement and not every single question (also no correclation within the topic block). I have to know in general if there is a correclation between the two topics.
I hope my issue is unterstandable.
Comments
This strikes me more as a confirmatory factor analysis?!
EJ
Hi EJ,
That might be one way to do this, but I think this might also be a feature request for a canonical correlation analysis?
I already have one open:
[Feature Request]: canonical correlation analysis, principal components regression, PLS regression and canonical powered partial least squares · Issue #2084 · jasp-stats/jasp-issues (github.com)
Best,
Tarandeep
Hmm yes, would also be good to have I guess.