Academic journal article International Electronic Journal of Elementary Education

Using First-Grade Teacher Ratings to Predict Third-Grade English Language Arts and Mathematics Achievement on a High-Stakes Statewide Assessment

Academic journal article International Electronic Journal of Elementary Education

Using First-Grade Teacher Ratings to Predict Third-Grade English Language Arts and Mathematics Achievement on a High-Stakes Statewide Assessment

Article excerpt

Abstract

Early childhood professional organizations support teachers as the best assessors of students' academic, social, emotional, and physical development. This study investigates the predictive nature of teacher ratings of first-grade students' performance on a standards-based curriculum-embedded performance assessment within the context of a state accountability system. The sample includes 4292 elementary school students cross-classified by 131 first-grade and 137 third-grade schools attended. This study uses extant statewide assessment data for students located in a state in the southeastern part of the United States. Controlling for student and school demographic variables in cross-classified random effects multilevel models, first-grade teacher ratings-as reflected by domain scores on a performance assessment-are found to positively and significantly correlate with students' third-grade academic achievement.

Keywords: Teacher ratings, Predictive validity, Curriculum-embedded Performance assessment

(ProQuest: ... denotes formula omitted.)

Introduction

Research on the use of teacher-based judgment measures (e.g., measures that use teacher ratings or rankings to assess students' knowledge or skills in specific academic content areas) to assess students' academic achievement in core academic areas (i.e., math, reading, science, or social studies) span over four decades. One of the most comprehensive scholarly reviews of the use of teacher-based judgment measures to assess student achievement was completed by Hoge and Coladarci (1989). The authors presented a thorough review of 16 empirical studies from 1962-1988. With these studies, they examined the association between concurrently administered direct and indirect teacher-based judgment measures in which teachers used ratings, rankings, grade equivalence, number correct, and item responses in reading, mathematics, social studies, and science, and norm-referenced measures of academic achievement. Overall, Hoge and Coladarci found a moderate to strong association (Mdn r= .66) between teachers'judgments of students' academic performance and their actual performance on standardized norm-referenced achievement tests.

Specific to the predictive value of teachers' ratings, several studies have investigated the longitudinal nature of teacher ratings in relation to students' performance on norm-referenced tests (e.g., Hecht & Greenfield, 2001; Meisels, Liaw, Dorfman, & Nelson, 1995; Quay & Steele, 1998; Stevenson, Parker, Wilkinson, Hegion, & Fish, 1976). Stevenson et al. (1976), followed a cohort of students from kindergarten through third grade and found moderate to strong correlations (rs ranged from .41 to .71) between teacher ratings (i.e., instructions, vocabulary, reflective, retention, learning, independence, or attention) and students' reading achievement, as measured by the Wide Range Achievement Test (Jastak, Bijou, & Jastak, 1965), across four time periods (i.e., prior to kindergarten and in the spring of grades 1,2, and 3), and moderate to strong correlations (rs ranged from .37 to .65) between teacher ratings (i.e., instructions, learning, vocabulary, retention, hardworking, independence, or reflective) and students' arithmetic achievement across the same four time periods.

Using a sample of kindergartners in three Michigan school districts to present predictive validity evidence for a performance assessment for young children, Meisels, Liaw, Dorfman, and Nelson (1995) reported high correlations between teacher ratings and students' performance on a norm-referenced measure within a one-year period. Specifically, the correlations between kindergarteners' total scores on the Work Sampling System (WSS; Meisels, Jablon, Marsden, Dichtelmiller, & Dorfman, 1994) checklists-an authentic assessment measure in which students are observed in their natural settings to determine the extent to which they can demonstrate proficiency on defined local, state, or national curriculum standards-and a total score on the Kindergarten Achievement Battery of the Woodcock-Johnson Psychoeducational Battery-Revised (WJ-R; Woodcock & Johnson, 1989), were . …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.