Psychometric g correlates with measurable biological variables including brain size, neural efficiency, and processing speed, proving it reflects real neurophysiological differences rather than statistical artifacts.