"Clearly the weight gain did not come from the exercise variable, but a 'missing' one (missing from the fridge)."

The reduction in calories from the fridge was the major variable causing the weight loss but for statistical purposes would there not be more correlation with the exercise variable if the number of calories burned during the time exercised were known? And might the correlations vary at each date in the time series?

And how would you account for the physiological changes due to exercise that caused you to become a more efficient calorie burning machine in your off-exercise time?

The statistics involved in weight loss in a complex, ever-changing organism seem quite daunting. Here for example is a recent study considering why lower caloric diets don't always have as strong an effect on overweight patients. 

Adam Robinson adds: 

Pitt raises an interesting point on systems thinking and caloric restriction.

It's well known that reducing calories induces your body to lower metabolism, counteracting the efficacy of the caloric restriction. An interesting variant is the observation many years ago that low intensity (aerobic) exercise burns more fat than does high intensity (anaerobic) exercise. Hence the "first order" (i.e., mistaken) conclusion that to lose body fat, high intensity exercise is better than low intensity exercise.

However, second order thinking (i.e., looking beyond the obvious) realized that this conclusion ignores, as Pitt points out, what happens the other 23-odd hours of the day and there the advantage goes to high-intensity exercise, which raises the body's metabolism longer throughout the day.

And it also ignores, ironically if you think about elusive mother nature, that because low intensity training burns more fat than does high intensity exercise, low-intensity devotees are quite possibly training their body to store fat to be used during future exercise periods! Again: advantage high-intensity exercise.

No direct applications to trading, but a reminder of the dangers in failing to look beyond one's immediate area of concern and failing to consider feedback loops. 





Speak your mind


Resources & Links