This report describes the construction and testing of two "psychological" learning models for the purpose of computer recognition of human speech over the telephone. One of the two models was found to be superior in all tests. A regression analysis yielded a 92.3% recognition rate for 14 subjects ranging in age from 6 to 13 years. Tests were done by having subjects respond to an elementary mathematics curriculum over their home telephones. The investigators conclude that machine-learning does occur and is amenable to theoretical analysis for the purpose of predicting the learning performance from the structure of the model. An attached appendix describes some preliminary experiments using audio for both input to and output from a computer-assisted instruction program. (Author)