Taking a test on content that has just been studied is known to enhance later retention of the material studied, but is testing more profitable than the same amount of time spent in review? High school students studied a brief history text, then either took a test on the passage, spent equivalent time reviewing the passage, or went on to an unrelated task. A retention test given 2 weeks later indicated that the test condition resulted in better retention than either the review or the control conditions. The effect was further shown to be content specific (in contrast to effects typically produced by questions inserted in text) and independent of item format. These results favor a greater use of testing in instruction.We wish to acknowledge the assistance of the Haverford Township School District, Delaware County, Pennsylvania, in the conduct of this study, especially the teachers who assisted directly: Mr. Bush, Mrs. McGarvey, Miss Harrison, Mr. Long, and their principal, Mr. Drukin.Requests for reprints should he sent to Ronald J. Nungester, The American College, 270 Rryn Mawr Ave.,

In a review of over 25 empirical investigations of effects of communicating behavioral objectives to students, several trends were apparent. Advance knowledge of behavioral objectives led to improved posttest performance in five of ten studies and to improved retention in two of three instances. Only two of seven studies found an interaction between knowledge of objectives and type of learning: in one case knowledge acquisition but not comprehension was facilitated, while in the other knowledge of objectives interfered on a problem-solving task, but not on a discrimination task. A third group of studies reported interactions between availability of objectives and reasoning ability, personality characteristics, and state anxiety. Finally, when coupled with complete learner control of the course, knowledge of objectives decreased learning time.

No abstract

No abstract

Learner control, which is contrary in spirit to the design philosophy underlying traditional CAI systems, is an important feature of most intelligent computer assisted instruction (ICAI) systems, as revealed through the learning dialogues they support. Such learner control is made possible by the natural language interface and the type of knowledge base incorporated in these systems. Such systems, it is argued, because they enable unplanned interactions with the learner and capitalize on student curiosity and motivation, can attain a much finer grain of individualization than traditional CAI systems can. The conceptual and affective requirements of instructional adaptation indicate the desirability of substantial learner control as well as the potential danger of overly strong tutorial guidance.