I was given a dataset at work, and I want to know how much 3 of the variables in it explain a fourth variable. I tried doing factor analysis, but calculating bartlett sphericity failed (I don't know what this means, I just know it's one of the steps). I think it may have failed because the dataset was too small? It's something like 400x50 points, although only 400x4 should really be interesting.

What is a good statistics/data science method to use here?

I tried basic stuff, no strong pearson correlations, nothing obvious when plotted, etc.

Thanks in advance