FURTHER EXPERIMENTS ON COMPUTER-AIDED LEARNING OF SOUND IDENTIFICATION.
Abstract
In experiments reported here the subject was able to choose freely, at any time, among the various instructional procedures, and to regulate the introduction of new materials. One of these experiments employed an electric typewriter as the means of communication between subject and computer; the other employed an oscilloscope display and light pen in an attempt to provide a more effective mechanism for response and feedback of results. Granting to the subject control of the training procedure led to no better final performance than that observed when the experimenter determined the course of the lesson, and the scope-and-light-pen device led to no better performance than the typewriter. Further analysis of the results provided detailed support of the earlier conclusions: for example, success was positively correlated with the proportion of time spent on simple exposure to the soundlabel pairs, and negatively correlated with the proportion of time spent in active responding, receiving feedback, and making second tries. Three variables related to streamlining of the task and temporal contiguity of sound and label account for nearly 100% of the variance. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 01, 1964
- Accession Number
- AD0602079
Entities
People
- Harry Rudloe
- John A. Swets
- Judith R. Harris
- Linda S. Mcelroy
Organizations
- BBN Technologies