Within-Class Cluster Grouping and Related Strategies: Brief Summaries

February 26th, 1998

Prepared by Joseph Renzulli and Harry Milne

The National Research Center on the Gifted and Talented
University of Connecticut, Storrs


Abrami, P. C., Chambers, B., D'Apollonia, S., Farrell, M., & De Simone, C. (1992). Group outcome: The relationship between group learning outcome, attributional style, academic achievement, and self concept. Contemporary Educational Psychology, 17, 201-210.
Purpose of Study
Examine impact of learning math under two cooperative group conditions (successful-unsuccessful) or students identified as (a) learned helpless or mastery oriented and (b) varying in math prior achievement (low, low-medium, high-medium, high). Dependent variables were mathematics achievement, causal attributions, and self concept.
Sample
6 classes of grade 7 math students (N=181) and 5 teachers (3 male; 2 female) from a suburban school in Montreal Quebec. 4 teachers taught 1 class and 1 teacher taught 2 classes using cooperative learning groups and STAD (Slavin, 1986).
Results
Nonequivalent pretest post-test design with statistical control for confounding factors. Group outcomes: one significant main effect. On average, students with less successful groups did significantly worse on the post test than students from successful groups, despite large class to class variation. Attributional style interactions: Helpless students in unsuccessful groups had lower posttest achievement scores than those in successful groups. Mastery oriented students were not affected by group outcome. Similar results for math achievement for those with low prior math achievement. Results did not apply to those with medium or high prior math achievement.
Conclusions
Learned helpless students and students low in prior math achievement from unsuccessful groups appear adversely affected by cooperative learning strategies such as STAD. Need to provide training and support. Need to further investigate class to class differences in implementation as well as teacher and student acceptance and understanding of strategies.


Armstrong, N. A. (1993). The effects of cooperative learning on gifted students in heterogeneous and homogeneous groups. (Doctoral Dissertation, Ball State University, 1993). Dissertation Abstracts International, 54-07A, 2457.
Purpose of Study
To investigate whether gifted students scored higher on reading achievement and self esteem ratings when they worked in heterogeneous, or homogeneous groups while involved in cooperating learning activities.
Sample
47 fourth grade students from 6 intact classrooms in four elementary schools in small midwestern school districts.
Results
Gates-MacGinitie Reading Tests used to test reading ability. The Coopersmith Self Esteem Inventory was used to measure self esteem.
Conclusions
Students achieve equally well in reading achievement and self esteem regardless of the grouping strategy used.


Berge, Z. L. (1990). Effects of group size, gender, and ability grouping on learning science processing skills using microcomputers. Journal of Research in Science Teaching, 27(8), 747-759.
Purpose of Study
To investigate one way to intervene in learning science processing skills to promote learning opportunities for all children using microcomputers as tools integrated into the science curriculum.
Sample
245 seventh and eighth grade students in 12 science classes in three schools in three different school districts selected on the basis of the number of computers available, and the teacher volunteering to be involved in the research. 12 classrooms were assigned randomly within the school buildings to one of three treatments: individual students, pairs of students and groups of four students.
Results
Integrated science processing skills were measured using The Test of Integrated Process Skills (TIPS) and The Test of Integrated Process Skills II (TIPS II) for Part I, and a researcher developed curriculum aligned test (Part II). Significant main effect ability gains were demonstrated. Part I posttest gain scores were highest for the low ability group and least for the high ability group. Part II gain scores resulted in an opposite trend. There were no gender related differences. There were no ability-treatment interaction effects. Instructional group size results were not consistent.
Conclusions
Students in teams of 2 to 4 in a microcomputer environment learn integrated science process skills as effectively as students working alone. There was no significant interaction between ability and group size. Procedures resulted in gender neutral achievement outcomes in science.


Bright, G. W., Harvey, J. G., & Montague-Wheeler, M. (1980). Achievement grouping with mathematics concept and skill games. Journal of Educational Research, 73, 265-269.
Purpose of Study
To examine differential effects of homogeneous and heterogeneous achievement grouping on mathematics learning through preinstructional concept games representing the concept of fairness and postinstructional skill games representing ordering of fractions.
Sample
164 students in eight seventh grade classes in one school in a northwest Chicago suburb: four classes were taught by each of two teachers. Two of each teacher's classes were randomly chosen to study probability; the remaining classes studied ordering of fractions.
Results
Gain scores were significant for all groups. No significant differences were found between heterogeneous and homogeneous groupings, or by gender.
Conclusions:
Suggests ways students are grouped to play preinstructional, concept games and postinstructional, skill games is not critical. Results indicate preinstructional games can be effective.


Dewar, J. (1963). Grouping for arithmetic instruction in sixth grade. Elementary School Journal, 63(5), 266-269.
Purpose of Study
To determine whether pupils who receive instruction in arithmetic at their level of achievement in a three group organization would show better achievement than pupils in a whole class organization, and to determine teacher and pupil attitudes to the methods of organization.
Sample
Eight sixth grade classrooms in a suburban, middle class area where teachers volunteered to take part in the study. Students were assigned to three groups in control and experimental conditions on the basis of Stanford Achievement Test in Arithmetic scores, and teacher judgment.
Results
Mean achievement gain scores indicated experimental groups one and three (3 ability groups for instruction, and traditional whole class organization) outperformed their control groups. There were no differences between experimental and control groups two. Attitudes of teachers indicated they saw better learning occurring in the former two experimental groups. Grouped students were seen to respond better to arithmetic. Pupils liked grouping.
Conclusions
Grouping for instruction appeared to be more effective in achieving learning. Teachers and students preferred the grouping procedure as it facilitated learning. Grouping procedure seemed to benefit high achieving and low achieving groups.


Felmlee, D., & Eder, D. (1983). Contextual effects in the classroom: The impact of ability groups on student attention. Sociology of Education, 56, 77-87.
Purpose of Study
To specify sources of heterogeneity on the individual and situational influences on the Rate of change from the state of attention to the state of inattention.
Sample
Classroom studied was a first grade classroom with a middle aged, female teacher and 23 students located in a middle class community in California. Students were primarily from middle class backgrounds. Students were assigned to four relatively equal size ability groups (two 'high', and two 'low') based on reading aptitude and teacher observation. 13 students in two higher groups (8 males; 5 females) and 10 students in the two low groups (5 male; 5 female).
Results
Reading aptitude level correlated significantly with low inattentiveness. Individual level variables: sex, maturity level, and past individual inattentiveness, while in predicted directions, were not significant. Group level variables: group ability level, reading length, and reading errors, have highly significant impact on the dependent variable. Group effects on inattention develop over time.
Conclusions
Group effects on inattention are much greater than individual effects. This emphasizes the strong impact of learning environments on students' behavior. Longer reading runs in less able groups correlate with increased inattentiveness. High group ability reduces incidence of inattentiveness and vice versa, especially over time, suggesting that these groups were not being exposed to equally positive learning environments.


Hallinan, M. T., & Sorensen, A. B. (1985). Ability grouping and student friendships. American Educational Research Journal, 22(4), 485-499.
Purpose of Study
To test the hypothesis that membership of the same ability group has a positive effect on the choice of a peer as close friend for elementary school children.
Sample
Taken from a large longitudinal data set on 1,477 students in 48 classes in 10 elementary schools in northern California. 658 (44.5%) were black, 697 (47.2%) were white, 75 (5.1%) were Asians, and 47 (3.2%) were Chicanos. Data collected from students in 104 reading ability groups in 32 classes over a school year.
Results
Significant results indicate membership of the same ability group over the school year promotes friendship formation among group members, especially in the earlier grades, for same gender and race, and where group structure remains stable. Larger groups (>8) having a greater influence than smaller groups (<8). There was a statistically significant decrease over time in the average proportion of students not in a friendship clique. Logit analysis indicated 50% dyads had the same gender; 75% were the same race; 37% were is same reading group. There was a positive effect of membership in the same group on best friend choice.
Conclusions
Analyses support the conclusion that membership in the same ability group increases the likelihood that students will become best friends, especially when groups are large. Network of best friends becomes denser over the year. Implies ability groups become more cohesive and attractive to members over time. Overlap between ability groups and friendship cliques increased over the school year. Teachers may be able to use ability grouping to assist students to develop social relationships.


Hoover, S. M., Sayler, M., & Feldhusen, J. F. (1993). Cluster grouping and elementary students at the elementary level. Roeper Review, 16(1), 13-15.
Purpose of Study
Assess extent to which cluster grouping is being used in Indiana as an option for gifted children and to determine how teachers, children, and parents react.
Sample
46 school districts applying for funding for gifted programs using some form of cluster grouping to the Indiana State educational agency.
Results
22 of 46 district (48%) responded to questionnaires. 75% indicated first or second year involvement. 67% teachers had no training in gifted education. 90% attended workshops, or conferences. 65% felt parents had positive attitudes. 49% reported regular classroom teachers were neutral. 48% teachers reported attitudes between cluster and non-cluster students were positive. 94% stressed thinking skills, 97% considered cluster grouping effective for the gifted. Cluster grouping benefits other students by having gifted students remain in the classroom.
Conclusions
Cluster grouping is relatively new in Indiana. Few schools have had cluster grouping for more than 3 years. Most use cluster grouping at the primary level. Generally positive regard expressed for cluster grouping from teachers and parents involved in cluster grouping.


Tanner, C. K., & Decotis, J. D. (1994). The effects of a continuous-progress, non-graded program on primary school students. ERS Spectrum, 12(3), 41-47.
Purpose of Study
To determine the effects of a nongraded, continuous progress program on five and six year old students attitudes toward school, as well as their readiness scores and their mastery of subject matter.
Sample
343 subjects in 14 elementary schools (175 girls and 168 boys) mostly from middle SES families. Experimental 159 CA 5 - 6 yrs (82 girls; 77 boys) enrolled in nongraded, continuous progress, cluster programs. C 184, (93 girls; 91 boys) in regular K and 1 in the same elementary schools.
Results
No differences on the Georgia Kindergarten Assessment Program (GKAP) scores. Report cards indicated higher ratings on categories reflecting initiative and independence for nongraded students. No affective differences were found by program. Affective differences found by gender SSA with girls having more favorable attitudes than boys. Readiness was not an issue in nongraded classes.
Conclusions
No significant differences in readiness scores. Report card scores indicated an advantage for the nongraded cluster program. Non-graded, cluster students surpassed controls in cognitive performance and exposure to higher order thinking skills. Advantages and disadvantages are listed.


Bierden, J. E. (1970). Behavioral objectives and flexible grouping in seventh grade mathematics. Journal for Research in Mathematics Education, 1, 207-217.
Purpose of Study
Utilizing intraclass grouping and behavioral objectives (Mager, 1962) to explore and develop a management plan for use in a mathematics classroom.
Sample
All 44 seventh grade students (two sections) in the Laboratory School of the School of Education, University of Michigan during 1967-1968. Mean IQ of 115, standard deviation of 22 as measured by the California Test of Mental Maturity.
Results
Measures of computational skills and knowledge of mathematical concepts increased significantly during the year. Test anxiety was reduced. There was a significant gain in attitudes towards mathematics, with the experimental method having a favorable influence. Varying degrees of flexible and relatively stable group membership combined with differentiated curriculum contributed to significant positive gains. Reactions of students to intra-class grouping and behavioral objectives were positive.
Conclusions
Flexible classroom management strategies and the use of behavioral objectives as the basis for a mathematics program were found to have positive outcomes in attainment, reduction of anxiety and attitudes of students to the program.


Hudgins, B. B. (1960). Effects of group experience on individual problem solving. Journal of Educational Psychology, 51(1), 37-42.
Purpose of Study
To extend knowledge related to problem solving behavior by examining if children working together in groups learn techniques of problem solving which they can apply later in similar transfer situations and to determine if the interaction contributes to successful group, as against individual problem solving.
Sample
128 fifth grade students selected in equal numbers of boys and girls from each of four public schools in the city of St, Louis. The 32 subjects in each classroom were matched by fours on the basis of their general ability and arithmetic test scores and randomly assigned to one of the four experimental groups.
Results
Phase 1: subjects working in groups gained higher scores than those working individually. Students working in groups without technique specification out-performed other experimental groups. Phase 2: results indicated there were no significant differences between groups in transfer of skills.
Conclusions
Results clearly oppose the hypothesis that problem solving must grow out of group interaction in the context of ad hoc groups. During phase 1 'leaders' often had their solutions accepted by the group unquestioningly. Others had to demonstrate the correctness of their solutions. A second pattern was more cooperative. Students solved problems independently and then compared answers. Phase 2 demonstrated that the mere knowledge of the principle will not insure transfer of training to new situations. Its general applicability must be realized.


Kamil, M. L., & Rauscher, W. C. (1990). Effects of grouping and difficulty of materials on reading achievement. National Reading Conference Yearbook, 39, 121-127.
Purpose of Study
The study examined the effects of two variables: small ability group instruction versus whole group istruction and difficulty of instructional materials compared to measured reading ability for students in the third, fourth and fifth grades.
Sample
869 students in 37 classrooms, solicited on a volunteer basis: 8 third grade, 12 fourth grade and 17 fifth grade classroom. 11used traditional instruction. 11 were whole group on level, and 14 whole group above level.
Results
No comprehension scores main effects were significant, but both covariates (group x comprehension and ability) were significant. Main effect for vocabulary ability, and two way interaction for treatment x vocabulary ability were significant.
Conclusions
Results suggest there was no essential difference between instruction patterned on a whole group model compared to the traditional homogeneous ability groups when the outcome measures are standardized achievement tests. The analysis also suggests that the discrepancy of the materials from the measured reading ability levels of the students accounts for a significant amount of variance in the scores. This suggests that a great deal of attention be given to the placement of students in materials, regardless of their ability, or the instructional organization of the class.


Knupfer, N. N. (1993). Logo and transfer of geometry knowledge: Evaluating the effects of student ability grouping. School Science and Mathematics, 93(7), 360-386.
Purpose of Study
To investigate the effects of student ability grouping on geometry learning transfer after a semester of instruction in a Logo environment.
Sample
Two classes of sixth grade students (N=50) of mixed ability, with the same teacher for computing classes, grouped according to their current ability level based on mathematics achievement, and randomly assigned to homogeneous (N=26: 8 each low, average and high ability,), or heterogeneous group (N=26 - 8 low, 8 average, and 10 high ability): situations.
Results
The score on the geometry posttest ANOVA revealed a significant difference in main effects on ability, but no significant differences between homogeneous and heterogeneous grouping patterns. A significant interaction between ability grouping pattern was found. Highest mean scores were made by high ability students regardless of homogeneous or heterogeneous grouping pattern. Lowest scores were achieved by low ability students grouped homogeneously, followed by average ability students grouped homogeneously. High students scored highest in homogeneous groups. Average and low ability students scored highest in heterogeneous groups.
Conclusions
During Logo exercises low ability students can benefit from heterogeneous grouping because they can receive more personal attention and more frequent in-depth explanations from student partners than from teachers. Average ability students appear to benefit from heterogeneous grouping, while high ability students appear to benefit from homogenous grouping, however these results are not clear. Logo exercises appear to have helped students to learn and transfer knowledge about particular geometric shapes. A number of further questions are raised. Results indicate such computer lessons should be made available to all students in cooperative learning situations, regardless of ability level.


Petty, M. C. (1953). Intraclass grouping in the elementary school (Bureau of Laboratory Schools Publication # 5313). Austin, TX: The University of Texas.
Purpose of Study
An exploratory study of the structural framework for intraclass organization for group guidance of children in elementary school classes. Three aspects are considered: (1) the relationship of the structure of the entire class to desirable practices in grouping children within the classroom, (2) the organization and guidance of intraclass groups, and (3) the role of instructional materials and equipment, and their implications for teaching arithmetic, reading and social studies.
Sample
Six classrooms, two sections each of grades one, three, and five. 100 boys and 92 girls, in 1951.
Results
Observations led to the conclusion that (ability based) group work was progressing on a very high level in all six classrooms. The children appeared to work with interest and pleasure in the intraclass groups.
Conclusions
A very comprehensive and basic, exploratory study that yields far more information that can be summarized here. An interesting report with an extensive reference list.


Sandby-Thomas, M. (1983). The organization of reading and pupil attainment. Journal of Research in Reading, 6(1), 29-40.
Purpose of Study
To examine how teachers' organization related to their aims and objectives, and to what students learned as well as the relationship between time, organization and the quality of instruction.
Sample
Six Belfast schools with working class backgrounds. From 400 students in 14 unstreamed, primary 2 classes (CA. 6), 34 students from individual classes were matched with 34 from group approach classes. In each sample of 34 students 12 were bright, 12 were average, and 10 were lows. Eight teachers grouped students of similar ability together for reading (group classes); six heard students read individually (individual classes).
Results
Diary record results indicated group teachers spend significantly longer 'hearing' reading. Individual class students spent more time with 'Breakthrough to Literacy'. Mean number of individual class students heard reading was 12; 20 for the group classes. More repetition for students taught in group classes. Mean reading ages higher for students taught in groups, with a significant difference only for 'bright' students. Significant differences found for comprehension, with group students higher than individual students. Low ability students in group classes tended to do better than those in individual classes but inability to score on the Neale Analysis of Reading Ability test resulted in insignificant results.
Conclusions
Reading instruction time was greater for group than for individual students. Four of eight group teachers prepared their instruction for the hearing reading situation. Six individual teachers did not prepare, but gave spontaneous instructions (prompting). Comprehension was emphasized by group teachers. Individual teachers could not find time to ask questions. Instruction was better prepared by group teachers. Pupil-teacher contact was longer and more frequent in the group than individual situation where there was also more repetition for weaker students. Group teachers provided more follow up and experienced fewer interruptions.


Slavin, R. E., & Karweit, N. L. (1985). Effects of whole class, ability grouped, and individualized instruction on mathematics achievement. American Educational Research Journal, 22(3), 351-367.
Purpose of Study
To investigate the mathematics achievement effects of three commonly proposed methods of dealing with student heterogeneity: individualizing instruction (Team Assisted Individualization TAI), within class ability grouping (Ability Grouped Active Teaching AGAT), and whole class instruction (Missouri Mathematics Program MMP) in different settings, (urban and rural) for students of different levels of prior achievement.
Sample
Experiment 1: 345 students (71% white, 26% black, and 3% Asian-American) in 15 grade 4-6 classes in a Wilmington, Delaware, school district formed as a consequence of an extensive desegregation plan.
Experiment 2: 480 students (91% white, 7% black, and 2% Asian-American) in 22 grade 3-5 classrooms in and around Hagerstown, Maryland.
Results
Experiment 1: All teachers implemented the major components of their methods. There were no pretest differences. For achievement, TAI and AGAT means were almost the same, and both were substantially higher than MMP. Attitude data for Liking of Math Class were significant with differences due to low scores for MMP classes. Both TAI and AGAT did not differ. On Self Concept in Math scale, significant results indicated TAI students scored much higher than AGAT and MMP, who did not differ.
Experiment 2: All teachers implemented the major components of their methods. Achievement data indicated significant pretest differences on Concepts and applications despite random assignment, but not for Computations. Post test scores were marginally significant for Computations, with AGAT and TAI not differing, but being superior to MMP. All three experimental conditions exceeded control. For Liking for Math Class significant results showed TAI ahead of AGAT and MMP. MMP exceeded control, but AGAT did not differ from control. There was no significance in results for Self Concept in Math.
Interaction effects for students at different achievement levels, in different treatments, and for sex and race in Experiments 1 and 2, were not found, contrary to expectations.
Conclusions
Commonality in findings across the two studies. Team Assisted Individualization (TAI) and Ability Grouped Active Teaching (AGAT) increased computational skills more then the Missouri Mathematics Program (MMP) and traditional whole-class instruction. No differences found between TAI and AGAT in either experiment. No statistical differences in Concepts and Applications. Achievement effects were main effects. No statistically significant interactions were found between treatment and absolute levels of prior achievement relative to class means, contradicting the expectation that effects of TAI and AGAT (designed to accommodate diverse achievement levels) would be most positive for students performing farthest from class means and in highly heterogeneous settings. Within-class ability grouping and methods that that include adapting instruction to diverse needs may produce gains in instructional effectiveness by providing appropriate levels of instruction, that may be partly offset by losses in instructional effectiveness due to the difficulty of managing multiple ability groups. If motivation and management problems can be overcome, this may be a particularly effective procedure. TAI continued to be used by teachers after the research program concluded.


Wallen, N. E., & Vowles, R. O. (1960). The effect of intraclass ability grouping on arithmetic achievement in the sixth grade. Journal of Educational Psychology, 51(3), 159-163.
Purpose of Study
To compare the achievement outcomes of intraclass ability grouping versus nongrouping procedures in arithmetic instruction.
Sample
112 students in four, sixth grade classes (N=25 in two classes, and N=31 in two classes) in two elementary schools in a new, middle class subdivision in Salt Lake City.
Results
No significant difference was found between grouping and nongrouping procedures, or Between Sequences or Between Schools. A significant teacher by method interaction was found. Interpretation being complicated by a significant school by semester interaction. Significant differences were found between students and between teachers.
Conclusions
Teacher personality had an influence on outcomes. Teacher sex interaction effect appeared with males preferring nongrouping and females preferring grouping. A teacher who can successfully individualize instruction may find grouping procedures a hindrance.


Webb, N. (1982). Group composition, group interaction, and achievement in cooperative small groups. Journal of Educational Psychology, 74(4), 475-484.
Purpose of Study
To verify the importance of group interaction for learning by investigating the relationship between student's interactions and their achievement in mixed ability and uniform ability small groups in junior high school mathematics classrooms.
Sample
96 students (40% minority members) from two average and two above average multiage/grade ability tracked Grades 7, 8, and 9 general mathematics classes in at a junior high school in the Los Angeles metropolitan area. Average mathematics and reading stanine scores of the Comprehensive Test of Basic Skills (CTBS, 1972) ranged from 3 to 9.
Results
Not receiving responses to questions had a significant, negative correlation with achievement. Introverted students were more likely than extroverted students to be ignored. Group composition and student personality had no direct effects on achievement, but they did have indirect effects on achievement through interaction in the group. Student ability had a direct effect on achievement. Most students indicated they enjoyed working in groups as "they could get help" and "it was more interesting than sitting in a stuffy class."
Conclusions
Interaction in the group mediated the effects of some group and student characteristics on achievement. Asking a question and receiving no answer was detrimental to achievement, and uniform ability groups produced more of this behavior than mixed ability groups did. This may be due in low ability groups to a greater incidence of off task behavior. Those in mixed ability groups stressed the importance of helping others. Lack of significance in achievement between uniform ability and mixed ability groups may be associated with the rule used for allocating students to groups, or to the inclusion of small groups as a factor in the analysis.