My last post about analysing my sleep data had plenty of caveats, but despite my caution I started to wonder whether I was taking an interest in the right variables.
