Introduction

The negative effects of reading problems are well documented (Harris & Sipay, 1990). There is evidence that reading disability is associated with social, economic, and psychological problems. There is little evidence, however, that efforts to correct reading problems through remedial reading programs or through special education placement have been very successful (Allington & McGill-Franzen, 1989; Hiebert & Taylor, 1994; Johnston & Allington, 1991; Karweit, Slavain, & Wasik, 1992-93; Kennedy, Birman, & Demaline, 1986; Rowan & Guthrie, 1989). Instead, there is evidence to suggest that children who encounter difficulty in learning to read fall further and further behind their achieving peers (Stanovich, 1986). Traditional approaches to dealing with reading problems, such as tracking and grade retention, do not help; indeed, they often appear to be detrimental to eventual student achievement (Shepard & Smith, 1989; McGill-Franzen & Allington, 1993).

In contrast, a growing body of evidence suggests that reading problems are preventable for the vast majority of students who encounter difficulty in learning to read, if these students receive extra support in the form of an early intervention program (Goldenberg, 1994; Hiebert & Taylor, 1994; Reynolds, 1991). In this paper, the term early intervention refers to early school intervention programs that are designed to prevent problems in literacy from developing rather than trying to correct a problem after it is established. For the most part, such programs have been used in first and second grades. Several of these programs have proven very effective when compared to conventional compensatory reading programs. For example, Hiebert, Colt, Catto, and Gury (1992) report that while 77 percent of the students in their early intervention project were reading at a primer level at the end of first grade, only 18 percent of a comparison group who participated in a traditional Title I program achieved that level of reading proficiency. While almost half (47 percent) of the students in the conventional Title I program remained nonreaders at the end of first grade, only 7 percent of the early intervention students were nonreaders. Thus, a growing body of evidence suggests that almost all reading problems are preventable.

A review of the research literature indicates that there are at least five early reading intervention programs that have documented effectiveness. This paper will only very briefly describe the individual programs and then will concentrate on the factors that seem characteristic of all or at least most of these successful intervention programs.

Two of the five programs, Success for All (Madden, Slavin, Karweit, Dolan, & Wasik, 1991; Slavin, Madden, Karweit, Livermon, & Dolan, 1990; Slavin, Madden, Karweit, Dolan, & Wasik, 1992), and the Winston-Salem Project (Cunningham, Hall, & Defee, 1991; Hall, Prevatte, & Cunningham, 1993), involve comprehensive reorganization of the entire classroom routine; all students in the grades in which the program is implemented are affected.

In the Winston-Salem Project all first and second grade reading/language arts instruction was reorganized around four major, thirty-minute blocks of activities: teacher-directed group reading activities, word learning activities, writing, and self-selected reading. In addition, in the school that served a very high proportion of at-risk students, an additional 45-minute block of time for very small-group instruction was included. During this small-group instruction time, students had additional opportunities to practice reading, writing, and word learning activities.

Though it has now been implemented in more than 85 schools, Success for All was first implemented in schools in major metropolitan areas that served, almost exclusively, children from low socioeconomic backgrounds who had few experiences with literacy. Major features of this schoolwide program included heterogeneous grouping for most of the day, cross-grade grouping according to reading level for 90 minutes of smaller group (15 to 20 students) reading instruction, and one-to-one tutoring for those students who needed extra support.

The Boulder Program (Hiebert, Colt, Catto, & Gury, 1992) and Reading Recovery (Clay, 1985; Clay, 1993; Clay, 1993a; Pinnell, 1989; Pinnell, Fried, & Eustice, 1990) are add-on, pull-out programs; they are added to whatever approach to language arts instruction is being used in a school. The Reading Recovery Program, which originated in New Zealand, has been widely implemented in the United States and in several other countries as well. It is exclusively a first grade, one-to-one tutoring program. Reading Recovery is also recognized for the extensiveness of its teacher training program, which is conducted over the course of a year with fully certified teachers.

The Boulder Program operated exclusively with the resources of a Title I program. The program worked originally with a pupil-teacher ratio of six students for each teacher. Through the use of carefully trained, supervised paraprofessionals who worked closely with certified, trained teachers, the ratio was reduced to three students for each teacher. Instruction was daily for twenty minutes.

The Early Intervention in Reading Program (EIR) (Taylor, Frye, Short, & Shearer, 1992; Taylor, Strait, & Medo, 1994) takes yet another approach. Regular first and second grade classroom teachers work for an extra twenty minutes with the five or six students who are encountering the greatest amount of difficulty in learning to read. Provision is also made for these students to practice their reading for an additional five or ten minutes each day by reading individually or in pairs to the teacher, a teacher's aide, a volunteer, etc. (For fuller descriptions of these programs, see Pikulski, 1994, or the citations in this paper.)