ACTIVITIES OF RESEARCH AND INSTRUCTION UNITS IN TWO MANITOWOC, WISCONSIN, ELEMENTARY SCHOOLS DURING THE 1966-67 SCHOOL YEAR ARE REVIEWED. RESULTS OF TWO CONTROLLED EXPERIMENTS CONDUCTED IN THE UNITS ARE REPORTED AND EVALUATED. IN ONE STUDY THE PERFORMANCE OF 82 THIRD GRADE PUPILS FROM ONE SCHOOL HOMOGENEOUSLY GROUPED FOR ARITHMETIC INSTRUCTION WAS COMPARED WITH THAT OF THIRD GRADE PUPILS IN A CLASS OF ALL ABILITY LEVELS. THE THREE HOMOGENEOUS GROUPS (HIGH, AVERAGE, AND LOW ABILITY) AND THE CONTROL GROUP (HETEROGENEOUS) WERE TAUGHT WITH THE SAME TEXT FOR 16 WEEKS. EACH TEACHER TAUGHT EACH OF THE FOUR GROUPS FOR A PERIOD OF 4 WEEKS. ALL PUPILS WERE TESTED ON THREE ARITHMETIC TESTS. RESULTS SHOWED THAT STUDENTS OF AVERAGE ABILITY-ACHIEVEMENT PERFORM BETTER IN HOMOGENEOUS GROUPS, STUDENTS OF LOW ABILITY-ACHIEVEMENT PERFORM BETTER IN HETEROGENEOUS GROUPS, AND HIGH ABILITY-ACHIEVEMENT STUDENTS PERFORM WELL UNDER EITHER GROUPING CONDITION. INDIVIDUALIZED AND TRADITIONAL SPELLING INSTRUCTION AT THE FOURTH GRADE LEVEL WERE CONTRASTED IN THE SECOND EXPERIMENT. FIFTY-SEVEN SUBJECTS WERE SEPARATED BY SEX AND RANKED BY SPELLING ABILITY. ONE-HALF OF EACH GROUP SERVED AS CONTROLS AND USED THE TRADITIONAL CURRICULUM. THE EXPERIMENTAL GROUP FOLLOWED A COMMERCIALLY DEVELOPED INDIVIDUALIZED PROGRAM. SPELLING LESSONS WERE GIVEN 15 MINUTES A DAY FOR 15 WEEKS. THE TEACHERS ALTERNATED CLASSES EVERY 3 WEEKS. EFFECTS OF THE INSTRUCTIONAL PROGRAMS WERE NOT SIGNIFICANTLY DIFFERENT FROM EACH OTHER. HOWEVER, STUDENTS PARTICIPATING IN THE EXPERIMENT GAINED 2 1/2 TIMES THE EXPECTED GAIN ON A STANDARDIZED SPELLING ACHIEVEMENT TEST. THE TEACHERS FELT THAT PARTICIPATION IN THE EXPERIMENT PROVIDED MOTIVATION FOR THE STUDENTS AND PROBABLY ACCOUNTS FOR THE ACHIEVEMENT GAINS. (AA/RS)