The achievement of the control students exceeded slightly the achievement of
students taught with computer management in two studies (Akkerhuis, 1974;
Coffman & Olsen, 1980), but the difference between groups in these studies was
trivial and nonsignificant. The achievement of CMI students was trivially higher
than that of control students in a study by Roberts (1982), but again the difference
between groups was nonsignificant. In a study of CMI by Nabors (1974),
however, the effect of CBE was positive and moderately high. The average ES in
the four implementations, however, was 0.07. The standard deviation of ES was
0. 196, and the standard error was 0.10.
The four studies provided little evidence for other positive effects of CMI.
Only the study by Akkerhuis (1974) examined noncognitive outcomes of instruction.
In Akkerhuis's study the average ES on attitude toward subject was -0.20,
and the average ES on attitude toward computers was -0.07. Both values are
small or trivial in size, and neither can be considered statistically significant.