THE DESIGN OF THIS STUDY WAS TO (1) DESCRIBE A WAY OF ANALYZING GRAPHIC OR PICTORIAL MATERIALS, (2) DESIGN SOME STIMULUS MATERIALS WITH THIS METHOD OF ANALYSIS IN MIND, AND (3) TEST SOME PREDICTIONS BASED ON IMPLICATIONS RELATED TO THE ANALYTICAL METHOD IN AN EXPERIMENTAL STUDY. A SET OF GRAPHIC MATERIALS WAS ANALYZED AND DESCRIBED WITHIN A STRUCTURE OF INFORMATION DEVELOPED BY WITTGENSTEIN (1921)--THE "PICTURE THEORY OF LANGUAGE." (FOR WITTGENSTEIN THE PRIMARY FUNCTION OF A PICTURE IS TO REPRESENT AND TO COMMUNICATE ITS REPRESENTATION. IF THE INTENDED REPRESENTATION AND A VIEWER'S INTERPRETATION OF A PICTURE AGREE, THEN THE PICTURE COMMUNICATES.) CONSIDERING THE WITTGENSTEIN CONCEPT, THE INVESTIGATOR DEVELOPED STIMULUS MATERIAL SETS, EACH CONSISTING OF A SAMPLE PICTURE TOGETHER WITH FOUR COMPARATIVE PICTURES. A TOTAL OF 240 SUBJECTS FROM 3 DIFFERENT GRADES PERFORMED "MATCHING-TO-SAMPLE" TASKS USING THESE STIMULUS SETS. TWO RESPONSE VARIABLES WERE STUDIED, SEPARATE SUBJECT GROUPS BEING SELECTED FOR EACH OF THE TWO - A RESPONSE TO SIMILARITY AND A RELATION RESPONSE. OTHER VARIABLES STUDIED WERE DIMENSIONALITY, SCHOOL GRADES, AND BLOCKS OF TRIALS. SEVERAL DEMONSTRATED INTERACTIONS AMONG THESE VARIABLES WERE EXAMINED, AND SOME THEORETICAL PREDICTIONS CONFIRMED. THE RESULTS PROVIDED DATA ON HOW TO USE GRAPHIC MATERIALS FOR EFFECTIVE TEACHING. (JH)