Dear Beatriz,
thanks for your answers!

Just one thing: Do i HAVE TO demean variables when running correlations.  I would normally not do this as I would think that, no matter the scale, the correlation analysis would reveal the same results.

The reason that I ask this is that I managed to get a bit strange results running one correlation analysis (see attached image) and I cannot understand what happened.

When running the same analysis but using Spearmans rather than Pearson correlation results seemed more like an result that I would expect from a shape analysis.

Any suggestions on what happened?

Thanks
Olof