ERIC Number: ED424251
Record Type: Non-Journal
Publication Date: 1998-Apr
Reference Count: N/A
Do Contextual Effects Bias Kentucky School District Accountability Index Scores? Occasional Research Paper No. 1.
Reeves, Edward B.
The system of high-stakes accountability in the Kentucky public schools raises the question of whether teachers and administrators should be held accountable if test scores are influenced by external factors over which educators have no control. This study investigates whether such external factors , or "contextual effects," bias the accountability index scores. The focus is on the Kentucky school district accountability index scores for the 1992-94 and 1994-96 biennial periods. District scores were chosen as the focus because more suitable measures of contextual effects are available. Three contextual effects are considered: (1) median household income in the district; (2) teen birth rate; and (3) rural-metropolitan differences among districts. Contextual effects might slow or promote a school district's ability to improve its scores, or they might only influence the scores within districts. A second set of analyses were done on the 4th, 8th, and 12th grade scores within each school district. Major findings are that contextual effects have little influence on the change in scores within districts. This supports the argument of high-stakes accountability advocates, who assert that since improvement is measured within the school or district, externally imposed advantages or disadvantages will not affect the results. Contextual effects have a large influence on differences in scores between districts. Between 30 and 40% of the variation in the scores is attributable to contextual effects. Median household income is the strongest of these effects, with rural-urban differences next. When contextual effects are controlled, rural school districts perform better than their accountability scores suggest and better than metropolitan districts. Urban schools perform less well than their scores suggest when contextual effects are controlled. Independent school districts tended to score higher than county school districts. Analyses from the individual grades generally support the overall analyses. An appendix lists scores and standardized residual for each district. (Contains 2 figures, 11 tables, and 20 references.) (SLD)
Publication Type: Reports - Research
Education Level: N/A
Authoring Institution: Morehead State Univ., KY. R & D Center for Students, Schools, and Communities.
Identifiers - Location: Kentucky