This article presents quality indicators for experimental and quasi-experimental studies for special education. These indicators are intended not only to evaluate the merits of a completed research report or article but also to serve as an organizer of critical issues for consideration in research. We believe these indicators can be used widely, from assisting in the development of research plans to evaluating proposals. In this article, the framework and rationale is explained by providing brief descriptions of each indicator. Finally, we suggest a standard for determining whether a practice may be considered evidence-based. It is our intent that this standard for evidenced-based practice and the indicators be reviewed, revised as needed, and adopted by the field of special education.

Early differences in family SES, child language production, and IQ were related to outcomes in early elementary school in the present prospective, 10-year longitudinal study. In a prior study of family interactional variables associated with language learning, major differences in parenting (i.e., time, attention, and talking) were found to be associated with differences in child productive vocabulary between 7 to 36 months of age, and child IQ, favoring higher-SES parents. Lower-SES children were exposed less often than higher-SES children to diverse vocabulary through their parents' attention and talking, and they were prohibited from talking more often. In the current study, 32 children involved in the earlier study were repeatedly assessed between 5 to 10 years of age, while in kindergarten through third grade. Results indicated that SES-related differences in child language prior to school were predictive of subsequent verbal ability, receptive and spoken language, and academic achievement assessed on standardized tests in kindergarten through grade 3. However, none of the predictor variables were related to direct measures of elementary schooling. When combined with a composite SES indicator, early child language production significantly increased the variance accounted for in the prediction of elementary language and academic competencies in each subsequent year in elementary school. Implications are discussed in terms of the stability of performance on language and academic performance measures of children who entered school with different early language learning experiences, and the need to consider early home- and school-based intervention designed to prevent or ameliorate these trends.

The results offer new insight into the landscape of the early language environment, with clinical implications for identification of children at-risk for impoverished language environments.

We conducted a longitudinal investigation of differences in classroom ecological arrangements and student behaviors (processes) on low-SES and high-SES elementary students' growth in academic achievement (products). An experimental, low-SES group received Cla.ssv.ide Peer Tutoring implemented by their teachers during each grade from first through fourth, while a low-SES control group and a high-SES comparison group received teacher-designed instruction. Results indicated that the experimental group and the comparison group, with distinctly different ecological arrangements and significantly higher levels of academic engagement, produced significantly greater product gains than did the control group. The implications of these findings are discussed.Research has established the importance of classroom processes to students* academic development (e.g., Brophy, 1986). Such processes include specific teacher-student interactions (e.g., Brophy, 1979), specific teacher behaviors and teaching functions (e.g., Rosenshine & Stevens, 1986), time allocated to academic tasks (e.g., Gettinger, 1985), and time students are engaged in specific academic behaviors (e.g., Greenwood, Delquadri, & Hall, 1984).Teaching practices known to orchestrate these processes include direct instruction (e.g., Becker, 1977;Becker & Gersten, 1982), mastery learning (e.g., Stallings & Stipek, 1986), cooperative learning strategies (e.g., Slavin, Madden, & Leavey, 1984), and peer tutoring (e.g., Greenwood, Carta, & Hall, 1988). In experimental evaluations, these practices have produced academic growth effect sizes from 0.25 to 2.00 standard deviations above control groups (e.g., Becker, 1977;Bloom, 1984;Walberg, 1986). These practices have also correlated positively with improvements in students* self-esteem (e.g., Becker, 1977) and peer relations (e.g., Warring, Johnson, Maruyama, & Johnson, 1985).

The purpose of this article is to discuss classwide peer tutoring as an effective instructional procedure. The article is organized into three major sections: (a) general principles of instruction, (b) description of classwide peer tutoring procedures, and (c) review of effectiveness data concerning classroom process (i.e., ecological and behavioral factors) and student achievement outcomes. It concludes with a discussion of the procedure and areas of future research and application.