A SERIES OF EXPERIMENTS WAS CONDUCTED TO TEST THE EFFECTIVENESS OF TELEVISED INSTRUCTION. THIS REPORT, THE SECOND IN A SERIES, EXAMINED THE EFFECTIVENESS OF ACTIVE STUDENT RESPONSE ON LEARNING DURING TELEVISED LESSON. PRINCIPLES OF PROGRAMING DERIVED FROM TEACHING-MACHINE RESEARCH AND APPLIED IN THIS STUDY INCLUDED (1) THE REDUCTION OF LESSON CONTENT TO SMALL STEPS AND THEIR SEQUENCING, (2) THE ENCOURAGEMENT OF ACTIVE STUDENT RESPONSE AT EACH STEP, FOLLOWED BY KNOWLEDGE OF RESULTS OR REINFORCEMENT, AND (3) THE PRELIMINARY TRYOUT OF LESSONS TO DETERMINE THE PROBABILITY OF ACTIVE RESPONSES AT EACH STEP. TWO LESSONS, ONE ON THE EFFECT OF HEAT AND ONE IN NUCLEAR REACTIONS, WERE EMPLOYED. IN EACH, A PROGRAMED VERSION WAS CONTRASTED WITH A CONVENTIONAL LECTURE. RESULTS TENDED GENERALLY, BUT NOT INVARIABLY, TO SUPPORT THE EFFECTIVENESS OF ACTIVE STUDENT RESPONSE TO PROGRAMED MATERIAL. THE GAIN IN EFFECTIVENESS WAS GREATER FOR HIGH IQ STUDENTS THAN FOR LOW IQ STUDENTS. THIS SUGGESTED THAT PROGRAMED TELEVISION LESSONS DIRECTED TOWARD THE AVERAGE STUDENT IN A GROUP MAY BE INADEQUATE TO ELICIT CORRECT RESPONSES. THE RESULTS INDICATED THE DIFFICULTY OF PROGRAMING TELEVISION LESSONS FOR GROUP INSTRUCTION WHEN THERE ARE WIDE INDIVIDUAL DIFFERENCES IN ABILITY. RELATED REPORTS ARE ED 003 646 THROUGH ED 003 653. (HB)