Race and ethnicity comprise only one set of indicators of student diversity. Other indicators show that an increasing number of students are growing up in single-parent and otherwise nontraditional family structures, more children are living in poverty, and more children are entering school with a primary language other than English and/or with limited English proficiency (ibid., pp. 25-28). Children also may exhibit diversity in less apparent ways. For example, there is some evidence that they may have different learning styles, which can have a considerable effect on their success in traditional learning environments (see Gardner, 1983).
In the context of formal schooling, being different has too often meant being deficient, and being deficient has meant "being at risk of academic failure." The compensatory education movement was founded in the 1960s on the assumption that many students, because of their minority and poverty status and their low academic achievement, are disadvantaged and should be provided with extra help and programs to "compensate" for those disadvantages. This "deficit model" has been criticized for rationalizing students' failure in terms of alleged deficiencies in their background -- a version of blaming the victim, which often serves to uncritically legitimize the existing school system (Baratz and Baratz, 1970, and Valentine, 1971).
Schools traditionally have responded to student diversity, poor academic performance, and dysfunctional behavior that are often the result of these differences with both organizational and programmatic approaches. Below we discuss instructional grouping and retention as two core organizational strategies that have been used, not without controversy, since the late 19th century. We also briefly consider special education and Chapter 1 pull- out programs as two more recent programmatic efforts to better educate students at risk.
Instructional grouping by ability is designed to enable teachers to most efficiently match content with students' apparent ability levels and learning paces. Both ability grouping and tracking have been severely criticized as methods for dealing with student diversity, however, because poor children and children of color are disproportionately represented in lower groups or tracks; there is evidence that lower level classes are often stigmatized and are likely to provide poor climates for learning and lower expectations for student achievement (Oakes, 1985, 1989, 1992; Slavin, 1989; Gamoran and Berends, 1987; Braddock, 1990). Students in the lower tracks often have restricted learning opportunities because the contents of their courses have been "dumbed down" and their teachers may be less experienced (in some cases, not even fully certified).
The relationship between different forms of instructional grouping and academic achievement is inconclusive, however. At the elementary level, Slavin (1987) synthesizes empirical evidence and shows that some forms of ability grouping do appear to be beneficial, especially when students are grouped for only one or two subjects while remaining in heterogeneous classes most of the day. He cites the Joplin Plan as an example of limited- ability grouping. Under the Joplin Plan, students are grouped by ability for reading only while remaining in heterogeneous groups for the rest of the day. Reading groups are cross- grade and assignments are frequently reassessed. In 11 of 14 studies reviewed by Slavin (1986), Joplin or Joplin-like plans showed consistently positive effects on student achievement. In contrast, he finds that the between-class, whole-day ability grouping at the elementary level has little or no positive effect on student learning.
At the high school level, students are often tracked into distinct academic, general, or vocational curricular streams. This has consequences not only in terms of the quality of education they receive but for peer-group formation, likelihood of graduation, and future educational and employment opportunities (Oakes, 1992; Braddock, 1990). Moreover, there is little evidence that students at the secondary level benefit academically from being in tracked classes (Slavin, 1990). In his analysis of several nationally representative datasets, Braddock (1990) finds that, in 1982, African-American and Hispanic students were significantly overrepresented in vocational tracks and significantly underrepresented in academic tracks when compared with white students. Moreover, studies show that completion of a vocational program often fails to significantly improve a student's employment prospects (Oakes, 1992). Finally, Gamoran and Berends (1987), in their review of a number of studies on tracking, report that the most consistent finding across studies is the effect of tracking on subsequent educational attainment: Students in academic tracks are much more likely to have college aspirations and to actually enroll in college than their nonacademic track peers.
As Oakes (1992) notes, "(S)chools serving predominantly low-income and minority students offer fewer advanced and more remedial courses in academic subjects, and they have smaller academic tracks and larger vocational programs (p. 567)." Restricted learning opportunities and persistent inequalities in the kind of education offered to disadvantaged students only puts those students at an even greater disadvantage, in school and in later life. While there is a great deal of controversy in this area, much of the evidence supports the conclusion that ability grouping and tracking as currently practiced do more harm than good for the overall educational achievement and attainment of low-performing, students at risk.
Retention. Like tracking, the practice of holding back students who fail to demonstrate required levels of achievement has been a common response to the challenge of educating low-achieving students. Also like tracking, the bulk of the research evidence shows that retention, as it is currently practiced in most schools, has few positive and mostly negative effects on student learning (see Shepard and Smith, 1989 for a collected review).
In a meta-analysis of 63 studies, Holmes (1989) finds that, on average across the vast majority of studies, retention produces negative results on such measures as academic achievement, personal adjustment, self-concept, and attendance. In the studies that showed positive results, retained students enjoyed unusually intensive remediation, were higher level retainees, and were often mainstreamed with their age peers for part of the day. Even in these studies, however, when compared with matched promoted students who also received extra help, retained students performed more poorly than did the promoted students.
Grissom and Shepard (1989) explore the effects of retention on dropping out of school and find that retained students are at greater risk of dropping out of school, controlling for poor academic achievement. These findings linking retention to dropping out are supported elsewhere (see, for example, Natriello, McDill and Pallas, 1990).
McPartland and Slavin (1990) point out that, as with tracking and ability grouping, retention might help improve the achievement of students at risk, but only if it is done in a "timely and effective" way (i.e., only holding back very young students who are less affected by the stigma of being retained, or only holding students back at certain key transition points in their school careers and providing them with high quality special programs if they have failed to master the skills required to advance).
Special education. Special education services have long been provided to students who have identified handicaps. Since the passage of Public Law 94-142 in 1975, school districts have provided services for handicapped students ranging from special schools to special classes within regular schools to various part-time placements. In these programs, students typically receive small group instruction from specially certified teachers.
In recent years, there has been a substantial increase in the number of students with mild academic handicaps who are receiving special education services. While the percentage of students categorized as physically disabled and mentally retarded stayed at about the same level from 1976 to 1989, the number of students categorized as learning disabled increased by more than 250 percent during the same period (National Center for Educational Statistics, 1990). According to Slavin (1989), almost 90 percent of this increase represents the entry into the special education system of low achievers who would not have been served in special education in the 1970s. Hence, he concludes, "special education has assumed a substantial burden in trying to meet the needs of students at risk of school failure..." in spite of the fact that "...research comparing students with mild academic handicaps in special education to similar students left in regular classrooms finds few benefits for this very expensive service" (Leinhardt & Pallay, 1982; Madden & Slavin, 1983).
Chapter 1 programs. The largest compensatory education program that provides extra help to impoverished students is the national Chapter 1 program. Chapter 1 began as Title I of the Elementary and Secondary Education Act of 1965 and continues today as the primary source of funding for a wide range of programs for students at risk. In the 1991-92 school year alone, Chapter 1 provided more than $6 billion to programs in 90 percent of public school districts serving approximately 5 million students nationwide (LeTendre, 1991; Anderson, 1992). Though some nonacademic services such as transportation, counseling, and health and nutrition programs are funded through Chapter 1, reading and mathematics instruction are the most commonly provided services (Anderson, 1992).
Most Chapter 1 programs follow one of five service delivery models: in-class, limited pull-out, replacement, add-on, or schoolwide. Because regulations require that Chapter 1 programs "supplement and not supplant" regular education services, and because, until recently, Chapter 1 funds had to be targeted only to eligible students, pull-out has been the strategy most widely used (Slavin, 1989; Birman, Orland and Jung et al, 1987; Natriello, McDill and Pallas, 1990). Under this model, students who are having difficulty in a particular subject typically are removed from their regular classrooms for 30 to 40 minutes per day to participate in subject-specific, small-group remedial instruction.
Stein, Leinhardt, and Bickel (1987) cite several disadvantages to the pull-out approach, including lack of coordination between what students learn in their regular classrooms and what they learn in the pull-out programs, disruption and wasted time spent traveling to and from pull-out classes, and diffuse responsibility for individual children. While Rowan and Guthrie (1989) find no evidence that pull-out programs at the elementary level were inferior to other models of Chapter 1 service delivery in terms of group size or amount of direct, interactive instruction, they also find that such programs often take away instructional time from other academic subjects.
Several evaluations of the Chapter 1 program conducted in the mid-1980s, including the longitudinal Sustaining Effects Study (Carter, 1984), concluded that Chapter 1 programs displayed modest positive effects on students' reading and math skills, but they were less effective for the most disadvantaged children. In any case, the gains overall did little to close the gap between disadvantaged students and their more advantaged peers. Moreover, students' progress was rarely sustained beyond 2 years after participation in the program (see the studies summarized in Natriello, McDill, and Pallas, 1990, pp. 72-78).