Saunak Sen (a statistician here at UTHSC) mentioned that the scree plot on the correlation matrix page appears to be incorrect. I noticed that the results differ from GN1's Scree Plot, but I haven't verified with Saunak yet whether GN1's results are correct (if they are, it can be used to verify GN2's Scree Plot).
See
for a demo on what it should look like using the GN API.