EXPERIMENTAL EXPLORATIONS IN PROGRAMMED INSTRUCTION AND OBJECTIVE TESTING MEASURES, REPORT OF THE 'VARIABLES INFLUENCING BEHAVIOR' PROJECT, PAPER 2.

AS ONE PHASE OF RESEARCH IN APPLIED ANTHROPOLOGY, YAQUI INDIAN AND MEXICAN MEN IN ARIZONA HAVE PARTICIPATED FOR THREE YEARS IN EXPERIMENTAL PROGRAMED COURSES IN BASIC ENGLISH AND ARITHMETIC. THE STUDENTS HAD PREVIOUSLY HAD AN AVERAGE OF FIVE YEARS' FORMAL SCHOOLING. A BATTERY OF STANDARD OBJECTIVE TESTS WAS GIVEN TO MEASURE IQ AND ABILITY TO READ AND COMPREHEND ENGLISH. A FIRST SERIES OF TESTS WAS ADMINISTERED WITH SPECIFIED TIME LIMITS OBSERVED. A SECOND SERIES WAS GIVEN ALSO WITH THE TIME LIMITS. THEN STUDENTS WERE ALLOWED TO COMPLETE THE TESTS. THE CONCESSION OF TIME TO WORK TO COMPLETION CHANGED IQ PERCENTILE RANKS FROM 11 OR 12 TO 66, 77.5 AND 63. VARIOUS TESTS OF READING, VOCABULARY, AND OTHER SKILLS YIELDED HIGHLY INCONSISTENT EVALUATIONS OF THE STUDENTS' ABILITIES. A THIRD SERIES OF TESTS WAS BASED ON A NOVEL EXPERIMENTAL APPROACH TO PROGRAMED LEARNING. AN ADJUSTING SCHEDULE OF REINFORCEMENT FOR MEETING TIME AND ERROR CRITERIA WAS PUT IN FORCE. THIS INVOLVED RAISING OR LOWERING REQUIREMENTS FOR REINFORCEMENT USING THE STUDENT'S OWN PERFORMANCE AS A MEASURE. RESULTS INDICATED THAT BOTH SPEED AND ACCURACY CAN BE CONTROLLED BY REINFORCEMENT CONTINGENCIES. (ALSO INCLUDED ARE SIX REFERENCES AND 32 TABLES.) (AUTHOR)