Effects of three variables in a teaching machine.

Conditions of Learning and Response Mode: an Experiment Using Programmed Learning Materials in an African Secondary School

“…The overall finding that different response modes did not produce significant differences in learning follows the general findings of Coulson and Silberman (1960), Evans (1960a), Challinor (1964), Hough (1962), but contrasts with the general tendency, reported by Grundin's (1969) survey of 47 studies for the constructed response to be superior when programs are over 100 frames in length, as the four programs in the present study taken together certainly were. The present findings follow the only other study of the effects of the response mode in Africa, that of Roebuck (1968Roebuck ( , 1970 who found over several schools no consistently significant differences attributable to the response mode.…”

“…The present findings follow the only other study of the effects of the response mode in Africa, that of Roebuck (1968Roebuck ( , 1970 who found over several schools no consistently significant differences attributable to the response mode. The finding that different response modes did not lead to significant differences in the scores for the four categories of learning contrasts with those studies where constructed responding produced significantly higher scores in particular areas or types of learning, Coulson and Silberman (1960), Fry (1960), Williams (1963), Suppes and Ginsberg (1962), but is similar to the findings of Evans (1960a, b) who found that multiple-choice responding led to equal scores on recall items (similar to verbal association items in the present study), similarly Hough (1962).…”

“…Such factors as time and the nature of the material to be learned constantly recur. Thus Coulson and Silberman (1960) and Evans (1960) found no overall difference on. post-test scores, but the multiple-choice response groups took significantly less time while, in the former case, the constructed response group scored significantly higher on test items requiring a constructed response.…”

Research on Programmed Learning

“…Though research with adults has usually found that linear and branching programmes on the same topic are equally effective (22), both Coulsen and Silberman (23) and Fry (24) have found constructed response programmes give higher scores on a test of recall though there is no difference when recognition tests are used.…”

The Effects of Linear and Branching Methods of Programmed Instruction on Learning and Retention of a Topic in Elementary Science

“…His conclusion was that the constructed response method was superior for recall. Coulson and Silberman (1960) gave branching and linear programs to junior college students and found the linear, constructed response group had a higher mean score on the constructed response section of the test, but there was no difference on the multiple-choice test. Roe (1962) found no significant difference between linear and branching methods except in time taken to learn.…”