Charles R. Greenwood, Joseph C. Delquadri, and R. Vance Hall (Juniper Gardens Children’s Project, University of Kansas), remind us that research studies have confirmed that classroom instructional techniques affect student achievement.
The amount of time students are engaged in specific academic tasks has been shown to influence student performance. Teaching techniques, including direct instruction, cooperative learning and peer tutoring, can increase the amount of time students spend on academics.
Cooperative learning and peer tutoring have also been linked to improvement in children’s self-esteem and peer relations. This research, however, has consisted mostly of short-term studies and Greenwood et al. state that there has been little research on the long-term benefits of these practices. They carried out a long-term project to investigate the benefits of classwide peer tutoring.
Greenwood et al. reported that in Chapter 1 schools (where there is a high proportion of students who live in poverty and are described as “at-risk” educationally), studies have revealed that less time is spent on academic tasks, including reading instruction. They hypothesized that differences in achievement between Chapter 1 schools and non-Chapter 1 schools may be due, at least in part, to differences in classroom teaching techniques and processes.
To test their theory, these researchers designed a long-term project to measure the effect a teaching strategy (peer tutoring) had on student academic behavior and achievement in Chapter 1 schools.
This tutoring program matched pairs of students who took turns tutoring each other. These pairs were assigned to one of two competitive teams. Pairs and teams were changed weekly.
Students earned points for their team by responding to assigned tasks which their tutor presented to them. The winning team was determined daily and weekly. The tutor/tutee roles were highly structured to ensure consistency, immediate error correction and rapid responding. Teachers prepared daily lessons to be taught and learned by each student.
The entire class participated simultaneously in tutoring and the teacher was free to supervise and monitor students. The peer tutoring method was designed to increase the proportion of instructional time in which students of diverse abilities are actively engaged in academic tasks.
Students followed from 1st to 4th grades
This classwide peer tutoring program was intended to follow students for 4 full years, from first through fourth grade. The project included both a “control group” of Chapter 1 (low socio-economic) school students who did not participate in the peer tutoring program, and the “experimental group” from similar Chapter 1 schools, who did. In addition, they followed a “comparison group” of students from high socio-economic schools. The control and comparison students received the standard instructional program. The researchers assessed the classroom processes of all three groups, as well as the achievement of each group. They also measured the strength of the peer-tutoring program itself, including the quality of its implementation and the efforts to maintain it over 4 years in 3 academic subjects: reading, math and language.
Peer tutoring for 90 minutes per day
All students took pretests and posttests each year. Five students in each class were randomly chosen to be observed twice a year (these observations were discontinued the fourth year due to a decrease in funding). Classes were selected from Chapter 1 and non-Chapter 1 schools in Kansas City, Kansas.
Each year, as the students in the study moved up another grade, project consultants trained teachers in the Classwide Peer Tutoring sytem.
Implementation of the tutoring called for beginning with one 30-minute spelling lesson. Once students and teachers felt comfortable and had developed proficiency with the technique, a mathematics and then a reading lesson were added to the program. Peer tutoring was intended to be used in the experimental classrooms for a total of 90 minutes each day. Greenwood and his colleagues provided not only initial training, but ongoing support in an effort to maintain the quality and strength of the treatment.
High attrition because of family mobility
The researchers report that the attrition rate among the students involved in the projects was more than 50% (68% in the experimental group) because of high family mobility and the closing of one school. The data also showed that, to a small extent, the experiment in peer tutoring may have influenced the teaching practices in the comparison and control groups.
It should be noted, too, that the design of the project did not allow researchers to determine which parts of the peer tutoring system (tutor/tutee interaction, error correction, competing teams) were most responsible for the achievement gains of the students.
Also, despite the effort to provide continuing support to teachers in the program, only partial implementation was achieved. For example, as mathematics became more complex, it was difficult for teachers to develop items which could be successfully taught or practiced in the tutoring format.
Additionally, the amount of effort involved in designing daily tutoring tasks in 3 subject areas and in monitoring pairs during these lessons was taxing for teachers. Some teachers also complained about the increased noise levels in their classrooms. Therefore, there was some reduction in the tutoring sessions, but this was not attributed to lack of student interest or motivation. Students were consistently very positive about their peer-tutoring sessions.
Peer-tutored students make gains
There were significant differences in achievement gain between the Chapter 1 students who participated in the tutoring program and those who did not. The experimental, peer-tutored classes made achievement gains comparable to those of the high socioeconomic comparison group. After four years in the program, the experimental group, whose members had been at risk for educational delays, exceeded or performed close to the national norm on standardized achievement tests in all three academic subjects.
The control group of “at-risk” students who received the standard instructional program plus Chapter 1 services (but no peer tutoring) remained below national norms. The experimental group exceeded the control group in all academic areas by .5 to 1.4 grade equivalents. These effects are equal to or greater than those achieved in other, shorter term studies.
Students in the experimental, peer tutoring program engaged in academic behavior (reading aloud, asking questions or discussing academics) more often than did the control group students. Peer tutored classes also spent less time on workbooks or behavior management.
Greenwood, Delquadri and Hall conclude that this study demonstrates that increased achievement gains can be obtained with low socioeconomic populations of “at-risk” students through sustained use of an effective instructional practice, specifically classwide peer tutoring. Even though peer tutoring was only partially implemented in this study, it produced significant increases in the amount of time students were actively engaged in academic tasks and, most importantly, it raised their achievement scores.
“Longitudinal Effects of Classwide Peer Tutoring” Journal of Educational Psychology September 1989 Volume 81 Number 3 pp. 371-383.
Published in ERN November/December 1989 Volume 2 Number 5