Abstract
The interaction between time of feedback and academic self-concept on performance in arithmetic skill was investigated using 236 fourth grade students. Using Tukey’s jackknife procedure, the regression lines of performance on perceived ability in spelling and reading were flatter for the group receiving 24-hour delay in knowledge of results than for the group receiving immediate knowledge of results. The flatter regression line for the 24 hour delay group was carried by students at the lower end of the perceived ability continuum having higher achievement scores than similar students in the other two groups. Thus, delayed feedback appeared to benefit students of low perceived ability.