TWO SETS OF STUDIES WERE CONDUCTED WHICH DEALT WITH PROBLEMS OF TEACHING MACHINE PROGRAMING AND TEACHING MACHINE DESIGN. ONE SET DEALT WITH THE VARIABLE OF STUDENT RESPONSE FOR THE PURPOSE OF DETERMINING WHETHER TEACHING MACHINE PROGRAMS SHOULD REQUIRE STUDENTS TO MAKE AN OVERT RESPONSE, AND IF SO, WHAT TYPE OF RESPONSE. A SECOND SET DEALT WITH VARIABLES AFFECTING THE DESIGN OF MACHINE TO DETERMINE WHETHER EXTERNAL PACING OF THE STUDENT WAS UNDESIRABLE, AND WHETHER THE TEACHING MACHINE POSSESSES MOTIVATIONAL PROPERTIES. RESULTS INDICATED THAT REQUIRING STUDENTS TO MAKE OVERT RESPONSES DOES NOT FACILITATE PROGRAMED LEARNING, THAT PACING THE LEARNER WILL NOT IMPAIR HIS PERFORMANCE IF CARE IS TAKEN TO ASCERTAIN OPTIMAL PACING RATES, THAT TEACHING MACHINES DO NOT PROVIDE A SIGNIFICANT ADVANTAGE OVER PROGRAMED TEXTBOOKS, AND THAT NEED TO REQUIRE STUDENTS TO MAKE ANY FORM OF CONSTRUCTED RESPONSE REMAINS TO BE ESTABLISHED. WHILE OVERT RESPONSES MAY NOT BE NECESSARY INSOFAR AS LEARNING IS CONCERNED, EXPERIMENTERS WERE NOT WILLING TO RECOMMEND THAT PROGRAMS BE WRITTEN TO ELIMINATE THE NEED FOR STUDENT RESPONSE. IT WAS FELT IMPORTANT TO RECOGNIZE THAT THE CONSTRUCTED RESPONSE (OVERT AND COVERT) MAY NOT PLAY A CENTRAL ROLE IN LEARNING, BUT IT COULD BE USEFUL IN A WIDE VARIETY OF PROGRAMED LEARNING SITUATIONS. IT WAS RECOMMENDED THAT EXPERIMENTAL WORK BE DONE TO DETERMINE METHODS OF PACING STUDENTS IN PROGRAMED LEARNING SITUATIONS, SINCE SUCH INFORMATION MAY PROVE TO BE USEFUL IN TRAINING. NUMEROUS TABLES ACCOMPANY THE TEXT.