PREVIOUS STUDIES REPORT THAT CHILDREN WHO ARE GOOD READERS DIFFER IN RESPONSE CONSENSUALITY FROM POOR READERS IN A WORK-ASSOCIATION TASK INVOLVING VERBAL OR PICTORIAL STIMULI. TO DETERMINE WHETHER THIS NONCONSENSUALITY EXISTS AMONG ADULTS FROM DIFFERENT READING LEVELS, 15 ADULTS WERE RANDOMLY CHOSEN FROM EACH OF THREE GROUPS--(1) AN ADULT BASIC GROUP READING BELOW THE THIRD-GRADE LEVEL, (2) A VOCATIONAL GROUP AT THE HIGH SCHOOL READING LEVEL, AND (3) A UNIVERSITY GROUP. TWENTY-FIVE BLACK AND WHITE LINE DRAWINGS OF COMMON OBJECTS WERE PRESENTED ON CARDS, AND THE SUBJECTS WERE ASKED TO WRITE THE FIRST ASSOCIATION THAT OCCURRED TO THEM. EACH SUBJECT'S RESPONSES WERE SCORED TWICE. FOR THE FIRST SCORING, THE UNIVERSITY GROUP SERVED AS CONTROL. THEN EACH SUBJECT'S RESPONSES WERE SCORED WITH HIS OWN GROUP AS CONTROL. A CONSENSUALITY SCORE CONSISTED OF THE TOTAL POINTS ACCUMULATED. RESULTS SHOWED THAT REGARDLESS OF THE SCORING SCHEME USED, THE UNIVERSITY GROUP'S MEAN CONSENSUALITY SCORE DIFFERED FROM THE SCORES OF OTHER GROUPS, AND THE VOCATIONAL GROUP'S SCORE DIFFERED FROM THAT OF THE ADULT BASIC GROUP. EVIDENTLY, CHILDREN AND ADULTS WHO ARE POOR READERS TEND TO RESPOND IDIOSYNCRATICALLY. THIS PAPER WILL BE PUBLISHED IN THE 1967 YEARBOOK OF THE NATIONAL READING CONFERENCE. (NS)