I ran CATREG and found that the R squared in model summary in the SPSS was inconsistent with the actual correlation between actual value and outcome. Have you encountered such problems? Is there a solution? Thank you for your time.
You could try posting your question to one of the following SPSS discussion forums:
http://spssx-discussion.165.s1.nabble.com/
https://listserv.uga.edu/scripts/wa-UGA.exe?INDEX -- search for SPSSXL
The first one does allow file uploads, so it might be the better choice. But people might be more willing to download a CSV data file rather than an Excel file. You should also show the code you used to generate your output. HTH.
Suggest you review catreg I have attached some helpful references. Please note that this is NOT ordinary least squares regression. Best wishes David Booth