Distinguishing between the g construct, measurement vehicles, and resulting scores reveals that practice effects and secular trends affect test-specific skills rather than underlying general ability.