Select the top performers after Test 1 and watch their Test 2 scores drift back toward the group average, even with no intervention. This is regression to the mean, not improvement or deterioration.
Parameters
Select Group
Group Statistics
No students in this group
Test 1 vs Test 2 Scores
Why does this happen?
Extreme scores on Test 1 include a component of luck. Top scorers were lucky; bottom scorers were unlucky. On Test 2, luck averages out, so both groups drift back toward the overall mean. No intervention is needed for this effect to occur.
Lower test-retest correlation (r) → stronger regression. At r = 1 all points fall on the identity line; at r = 0 there is no relationship at all.