Rock climbers, in my experience, are not especially given to discussing statistical factor loadings. For climbers, figuring out the loads on each element in the rope system has, shall we say, more grave and immediate consequences.
Straddling the worlds of p-values and Yosemite Decimal System ratings, as I do, and being given to a predilection for imposing my geekiness on others, I fairly squeaked with joy when I found today a rock-climbing illustration of a statistical concept. It is (tum ta tum!) the scree test. It's a graphical way of deciding which factors to retain following a principal components analysis. You plot your eigenvalues on a line plot, which ends up looking at the left like the steep mountains we climbers so enjoy surmounting, and on the right like the more gently sloping (but still arduous to traverse) scree fields (piles of jagged rocks just right for ankle-twisting) you typically find at the base of a big, juicy rock face. You then dump the factors that form the scree field because they're, well, scree, and nobody likes it.
Pictured here is the one person in the world I can think of who'd be likely (p <.05) to be just as excited about this as I am.