Correlation is not causation? Do we actually need experimental units to undergo a controlled experiment with random assignments of treatments to confirm what is self-evident about determinants of aggression? Ok maybe you do. So take 30 people and feed them. Then take another 30 and starve them. Report back to us which…