A COMPARISON OF THE EFFECTS OF MULTIPLE‐CHOICE AND CONSTRUCTED RESPONSE PRE‐TESTS IN PROGRAMMED INSTRUCTION

Following some previous research, the aim of this experiment was to test the hypothesis that taking a constructed‐response pre‐test would improve a subject's post‐test score in programmed instruction by increasing curiosity but that taking a multiple‐choice pre‐test would not have this effect because it would concomitantly increase incorrect learning. Three groups of subjects (aged 10 to 12) were compared: group 1 took a multiple‐choice pre‐test; group 2 took a constructed response pre‐test; group 3 took no pretest at all. This was carried out at two different ability levels (above average and below average). All subjects were given a linear program followed by a post‐test which was the same for all groups at each ability level. There were no significant differences between the post‐test scores of any of the groups at either level, and therefore the hypothesis was not supported. It is suggested in the light of this result and that of previous experiments that the age factor may be the crucial one...