THE EFFECTS OF CERTAIN RESPONSE CHARACTERISTICS IN PROGRAMED INSTRUCTION ON ERRORS, RATE OF LEARNING, AND RETENTION.

TO INVESTIGATE THE ROLE OF MULTIPLE RESPONSE ALTERNATIVES AS CONSTRAINTS ON THE MEANING OF THE CORRECT RESPONSE, 200 SUBJECTS IN AN INTRODUCTORY EDUCATIONAL PSYCHOLOGY COURSE WERE RANDOMLY ASSIGNED TO 8 TREATMENT GROUPS VARYING ITEM DIFFICULTY, RESPONSE AND SELF-CORRECTION (OVERT V. COVERT), PLUS ONE GROUP FOR USUAL LINEAR PROGRAMING INSTRUCTION. ANALYSIS OF VARIANCE ON IMMEDIATE AND DELAYED ACHIEVEMENT POST-TESTING, USING HALF MULTIPLE CHOICE AND HALF COMPLETION ITEMS SHOWED THE FOLLOWING. EASY ITEMS, CONSTRUCTED RESPONSES AND ERROR CORRECTION RESULTED IN A REDUCED ERROR RATE. DIFFICULT FOILS AND ERROR CORRECTION RESULTED IN SUPERIOR PERFORMANCE ON THE IMMEDIATE COMPLETION TEST, BUT NO FACTORS RESULTED IN DIFFERENTAL PERFORMANCE ON THE DELAYED COMPLETION TEST. THE LINEAR PROGRAM WAS MORE EFFECTIVE FOR ALL CONDITIONS EXCEPT DIFFICULT FOILS AND CORRECTION PROCEDURE. DIFFICULT ITEMS AFFECTED PERFORMANCE ONLY ON THE DELAYED MULTIPLE CHOICE TEST. EASY ITEMS AND THE LINEAR PROGRAM PROMOTED IDENTICAL RESPONSE ERROR REPETITIONS FROM THE IMMEDIATE TO THE DELAYED TEST. RESULTS SUGGEST THAT DIFFICULT, PLAUSIBLE ITEMS COUPLED WITH A CORRECTION PROCEDURE, MAY BE THE MOST EFFECTIVE ADAPTATION OF THE LINEAR PROGRAM. ALL SUBJECTS WHO TOOK THIS PROGRAM OUT-PERFORMED THOSE EXPOSED TO THE LINEAR PROGRAM ON EVERY CRITERION MEASURE. (LH)