Effects of Adjunct Questions, Pretesting, and Degree of Student Supervision on Learning from an Instructional Text

Fifty-eight students undertook self-paced study of a 23, 000-word text. The 2x2x2 design involved: (1) adjunct questions versus no adjunct questions, (2) a pretest versus no pretest, and (3) supervised study versus unsupervised study. The posttest included the pretest items and adjunct questions along with new items. A retention test came 7 to 13 days later. Neither pretesting nor supervision had any influence on achievement. Adjunct questions did improve performance, but only with respect to that subtest of the posttest comprised of the adjunct questions themselves. Thus, adjunct questions failed to produce the general beneficial effects on learning observed elsewhere by others. This appears attributable to the experimental requirement of other studies that Ss read the instructional material only once.