Differences in Instructional Quality Between High- and Low-Performing Schools

• In Event: Effective High Schools: Insight From the National Center on Scaling Up Effective Schools

Abstract

We hypothesized that higher value added schools provide higher quality instruction than lower value added schools, with instruction defined as the strategies and assignments teachers use to enact their curriculum. Research suggests that these strategies are most effective when students use knowledge to solve problems, engage in simulations, or apply knowledge to new contexts (McLaughlin & Talbert, 1993; Wenglinsky, 2002, 2004).

We investigated quality of instruction in two ways: observation of classroom instruction and coding of interviews with school administrators, teachers, and students regarding how schools are explicitly or implicitly organizing for improving instruction. In each of the four case study schools, we observed instruction in all English Language Arts (ELA), mathematics, and science classrooms in 10th grade---310 different class periods taught be 73 different teachers. This allowed us to explore instructional quality across tracks in ELA (e.g., remedial, general, college prep) and course sequences in science and mathematics. For cost and feasibility reasons, we chose grade 10, as it is the last common year in which the district requires students to take standardized exams in mathematics and ELA.

Instruction was “live coded” using the Classroom Assessment Scoring System-Secondary (CLASS-S), an observational tool for observing and assessing the quality of teacher-student interactions in classrooms. Based on development theory and research suggesting that interactions between students and adults are the primary mechanism of student development and learning (Greenberg, Domitrovich, & Bumbarger, 2001; Hamre & Pianta, 2006; Morrison & Connor, 2002; Pianta, 2006; Rutter & Maughan, 2002), the CLASS-S focuses on interactions across three domains: Emotional Support, Classroom Organization, and Instructional Support. The K-5 version of the CLASS has been well-validated, with classrooms that obtain higher CLASS scores having students who make greater academic and social progress during the school year than students in low-scoring CLASS classrooms. Studies have shown a high degree of reliability across coders as the result of required, intensive training. All coders in our study were trained by the developers of the CLASS-S.

Preliminary analyses suggest that differences among school means are found primarily in the Emotional Support (e.g., Regard for Adolescent Perspective, F(3,284)=2.74, p<.050) and Classroom Organization domains (e.g., Behavior Management, F(3,284)=3.89, p<.01; Productivity, F(3,284)=5.08, p<.01). Primary differences are between school A, a neighborhood school with an opt-in requirement and school D, a lower value-added school. We also saw evidence that the difference between high and low tracks in school A was narrower than in the other case study schools.

Interviews with principals, teachers, department heads and various support personnel were coded for evidence of quality instruction by a team of researchers. They first coded transcripts individually according to our theoretical framework based on the CLASS, while allowing other domains to emerge, then arbitrated codes to reach consensus and establish inter-rater reliability. Preliminary analyses of interview data regarding how schools organize for quality instruction suggest key elements which distinguish effective schools from those that are ineffective include a focus on maximizing instructional time and instructional flexibility allowing teachers to reteach material their students have not grasped.

Authors

• Thomas M. Smith, Vanderbilt University

• Bruce Vineyard, Florida State University

• Courtney Preston, Vanderbilt University

• Karin Katterfeld, Westat

• Katherine Taylor Haynes, Vanderbilt University