Looking at reviews is essential. In describing the research on nongraded elementary schools, the first thing I need to say is that the vast majority of the research has centered upon the primary grades (K-3). Thus, there is much less research to use in judging the effect of such programs for grades 4-6. Also, we are fortunate that several educationists have reviewed the research on nongraded schools. Reviews are helpful here because the quality of the research in this area is quite variable and it is published in diverse locations (manuscripts of meeting presentations, theses, magazines and journals). The research studies may involve use of small numbers of students so that the statistical certainty of conclusions drawn from a single study is low. Therefore, by looking at trends of results of multiple research studies, which have been selected for their good quality, one can try to answer the question, "How likely will it be that a nongraded program will produce a positive, negative, or neutral effect upon student achievement?"

The reviews give conflicting conclusions. Even the reviews of research studies of the nongraded classroom have reached varying conclusions about the usefulness of this approach. After presenting an overview of the various reviews of research studies, Ellis and Fouts (1) stated that the majority of the reviews could not find conclusive evidence that the nongraded approach changed student achievement while a few concluded that children may do slightly better in nongraded classrooms. Ellis and Fouts pointed out that a number of reviews have supported the idea that "multi-age grouping appears to offer some advantages in affective growth, particularly self-esteem, aspirations, feelings of success, and perceptions of parental approval." I will go on record as saying that this may indeed be true, although some of the data used to support this premise are vague. However, I don't think a particular nongraded program should be used unless one can expect it to benefit student achievement in reading, math, and other "cognitive" subjects.

Since 1990, three reviews of good quality have appeared which look at research reports of studies contrasting student achievement in nongraded classes with achievement of students in traditional graded classes. These reviews screened research reports from approximately 1960 through about 1985. The authors of each review used slightly different techniques for choosing reports viewed as being of good quality. They also used different methods to analyze the research reports they chose to review. For example, Anderson and Pavan (3,4), using a score card kind of analysis, concluded that studies showing a positive effect of nongraded school on student achievement far outnumbered significant negative ones. Thus, Pavan, who is a known advocate of the nongraded approach, concluded that nongraded programs are likely to benefit most schools.

By contrast, the other two major reviews of nongraded schooling, which used a method to analyze the research different from Anderson and Pavan, reached different conclusions. The authors of both of these reviews calculated effect sizes for the mean (average) achievement test scores of students enrolled in each research project. To calculate an effect size they calculated the difference between the mean test scores for the students in nongraded and graded programs for each study and divided this by the graded program's standard deviation from the mean. The idea is to get some idea about how much the nongraded program influenced achievement, taking into account the fact that there would be a significant variation in how well students would do on an achievement test whether they were in a nongraded or graded program. A positive effect size would indicate achievement appeared to be higher for students placed in nongraded classrooms, and the larger the positive effect size observed, the more likely the positive effect would occur if a nongraded program were used in multiple schools for many students. On the other hand a negative effect size would indicate placement in a nongraded classroom tended to decrease student achievement. An effect size of zero would mean there was no discernible difference between achievement of students in the nongraded and graded classrooms.

Using effect sizes, Veenman (5) reviewed what he thought were the best studies available and concluded that cognitive achievement for reading, mathematics, and language was not different for students in nongraded compared to graded classrooms. The median (most frequently observed) overall effect size he observed was -0.03, which was very close to zero (no effect). The effect size for noncognitive parameters (self-esteem, personal adjustment, social adjustment, and motivation) also did not show that the nongraded approach was clearly beneficial compared to the traditional graded approach (median effect size +0.15). Veenman was a "lumper" rather than a "splitter" in looking at studies comparing nongraded and graded schooling. He did not look at the effect of particular types of nongraded programs on student achievement. Obviously, if he reviewed studies that represented a mixture of program types, the beneficial effect of a particular program type might be lost in the "background noise" of the remaining studies reviewed. Ignoring this possible criticism of his work, it is still useful to list the factors which he believed could contribute to his inability to see differences in student achievement between the nongraded and graded programs. Some of these give insight into some of the potential pitfalls which could occur when a school decides to undertake a nongraded approach. The factors that he speculated could have led to an inability to demonstrate improved achievement using a nongraded approach were:

• The grouping strategy chosen for schooling may not be important. Successful learning might be much less dependent upon organizational strategies than upon the quality of instructional practices used. This is an interesting proposition. Please keep it in mind as you read about the conclusions reached by Gutiérrez and Slaven, which I will discuss a little later.
• There might have been bias in the selection of students placed in the nongraded and graded classrooms. The differing characteristics of the students then canceled any effects produced by the grouping strategies used.
• Teachers in a nongraded classroom may have been ill-prepared to teach two or more grades at the same time and may not have had teaching materials that were adequately suited to the nongraded classroom.
• There may be a teacher "burn out" factor. He notes that teachers in nongraded classes who were surveyed indicated that nongraded classes involved a greater workload, required more preparation time, and demanded better classroom management skills.

1. Joplin-like plans (9 studies)--nongrading occured for only one subject (8 in reading and 1 in math); otherwise children remained in graded classrooms.
2. Comprehensive (14 studies)--nongraded programs emphasized continuous progress and flexible, multi-age grouping, but did not emphasize individualized instruction.
3. Individualized (12 studies)--these programs emphasized individualized instruction, learning stations, learning packages, programmed instruction, and/or tutoring.
4. IGE (10 studies)--the characteristics of these programs were described in detail earlier.
5. Unspecified (12 studies)--these studies failed to state what was actually implemented in the nongraded programs.

Table 2
Summary of Effects by Type of Nongraded Plan
[from Gutiérrez and Slavin (6)]

Type of Program	Total Studies	Signif. Positive	Nonsig. Positive	No Diff.	Nonsig. Negative	Signif. Negative	Median Effect Size
Joplin-like	9	4	2	1	1	1	+.46(7)
Comprehensive	14	8	2	1	3	0	+.34(9)
Individualized	12	2	6	2	2	0	+.02(9)
IGE	10	4	3	2	1	0	+.11(6)
Unspecified	12	2	3	1	4	2	+.01(6)

Their main conclusions were as follows:

[The] positive effects of nongraded organization are most consistent and strongest when the program focuses on the vertical organization of the school and when nongrading is used as a grouping strategy but not as a framework for individualized instruction. Four categories of nongraded programs were examined, in addition to one group of studies in which the nature of the nongraded program could not be determined. Studies in two of these categories clearly supported the nongraded plans. These are the Joplin-like programs, in which students are grouped across age lines in just one subject (usually reading), and the comprehensive programs, which involve cross-age grouping in many subjects but still rely on teacher-directed instruction. The median effect sizes for studies in these categories were clearly positive (+.46 for Joplin-like programs, +.34 for comprehensive), and the best designed evaluations were the ones most likely to show the positive effects. In contrast, nongraded programs that incorporated a great deal of individualized instruction (and correspondingly less teacher-directed instruction), including Individually Guided Education (IGE), were less consistently associated with achievement gain. This is not to say that these approaches reduce student achievement; their effects are very inconsistent, generally neither helping nor hurting student achievement, with more studies finding positive than negative effects (especially in the case of IGE). Poorly described nongraded programs also had median effect sizes near zero, perhaps because experimental and control groups may not have differed in anything essential except label. What accounts for the relatively consistent positive effects of the Joplin-like and comprehensive nongraded plans and the less consistent effects of programs incorporating individualization? At this remove of time from the flowering of the nongraded ideal, one can only speculate, but there are many more recent developments in educational research that suggest some possibilities. The most obvious reason that incorporating a great deal of individualization might have reduced the effectiveness of the nongraded elementary school is suggested by research on individualized instruction itself, which has generally failed to support this innovation [4 references cited]. Correlational evidence from process-product studies of more and less effective teachers has consistently found that student learning is enhanced by direct instruction from teachers, as contrasted with extensive reliance on individualization, seatwork, and written materials [reference]. Further, to the degree that the nongraded elementary school came to resemble the open school, the research finding few achievement benefits to this approach [reference] takes on increased relevance.... If the effectiveness of nongraded organization is due to increased direct instruction delivered at students' precise instructional level, then it is easy to see how a move to greater individualization would undermine these effects. Individualized instruction, learning stations, learning activities, and other individualized or small group activities reduce direct instruction time with little corresponding increase in appropriateness of instruction to individual needs (in comparison to the simpler cross-age plans)(6)[emphasis added].

Most of the studies reviewed above took place before 1985. What relevance do they have when parents must judge whether or not a nongraded elementary school program is a good idea for their child? First of all, if the nongraded approach being offered for their child is a simple one, similar to the Joplin plan, the parents can have confidence that the program may very well enhance their child's achievement. However, if a school is using a comprehensive nongraded program or an IGE program, parents should be more cautious. If the school's personnel start talking about every child having an individualized learning plan to fit his or her learning style, parents ought to be concerned about whether that program is going to improve the child's achievement.

Today parents also need to look at any proposed nongraded program to determine if it differs from those that were used between 1965 and 1985. If the nongraded program is bringing in other "innovations," parents have a right to be quite concerned.

For example what would happen if one took an individualized or IGE nongraded program and simultaneously began to use an OBE approach featuring use of vague outcomes to drive the curriculum and formal, detailed portfolios to demonstrate students' knowledge, skills, or behaviors are aligned with the outcomes? Obviously, the answer to this question will be a speculation. In my opinion, a program that goes this route is likely to have adverse effects on student achievement. The curriculum will inevitably de-emphasize teacher-directed instructional methods which have been shown to work. The curriculum will lack a clear focus on teaching basic skills in reading and mathematics. The use of formal portfolios would burden teachers who are already burdened with the extra work involved in managing a multi-age classroom. All of these factors will adversely effect student learning. Fortunately, at this point in time, it appears that the OBE movement's influence may have peaked and it is in decline. Thus, it is less likely that the more extreme practices associated with this approach to schooling will be incorporated into nongraded classrooms.

Perhaps the most important lessons to be learned from past implementations of the nongraded classroom are these:

1. Most of the experience with the nongraded approach to schooling has been with primary grade children (grades K-3). Thus, we don't know as much about how well they work beyond this point in elementary school.

2. "Keep it simple." The more subject areas which are treated in a nongraded manner and the more innovations which are introduced into a nongraded program, the more difficult it becomes to predict the resultant product will be successful in raising student achievement.

3. When nongraded grouping is done for a single subject such as reading with the expressed intention of enhancing teacher-directed instruction, the likelihood of improved student achievement in the school is increased. On the other hand, if a nongraded program is established mainly to provide a framework for individualized instruction, one cannot have as much certainty that the school will be successful in improving student achievement.

1. Ellis, A.K. and J.T. Fouts (1994). Research on school restructuring. Princeton Junction, N.J.: Eye on Education. pp. 147-161.

2. Klausmeier, H.J., M.R. Quilling, J.S. Sorenson, R.S. Way, and G.R. Glasrud (1971). Individually guided education and the multiunit elementary school: guidelines for implementation. Madison: University of Wisconsin, Wisconsin Research and Development Center for Cognitive Learning.

3. Anderson, R.H. and B.N. Pavan (1993). Nongradedness: helping it to happen. Lancaster, PA.: Technomic.

4. Pavan, B.N. (1992). The benefits of nongraded schools. Educational Leadership 50 (2): 22-25.

5. Veenman, S. (1995). Cognitive and noncognitive effects of multigrade and multi-age classes: a best evidence synthesis. Review of Educational Research 65(4):319-381.

6. Gutiérrez, R. and R.E. Slavin (1992). Achievement effects of the nongraded elementary school: a best evidence synthesis. Review of EducationalResearch 62(4):333-376.

Return to PRESS Home Page