Factors explaining success and lack of success on the Michigan Educational Assessment Program science assessment.

Abstract

The fundamental purpose of this study was to determine how the science programs of schools that demonstrated high achievement on the Michigan Educational Assessment Program (MEAP) science test compared with those of schools which obtained low scores on that test. Two research questions provided a context for this inquiry: Compared to non-school variables such as socioeconomic level, how important are science program variables in explaining achievement on the tenth grade MEAP science test? What is the relationship between the science section and the reading and mathematics sections of the test? One hundred three middle school principals were mailed surveys which asked about the school community and the school science program. Variables included in the instrument were fashioned from the criteria for excellence described in the National Science Teachers Association publication Focus on Excellence: Middle School/Junior High Science. In addition, the Michigan Department of Education supplied data about schools and school districts. Socioeconomic level was found to be a strong predictor of student achievement in science whether it is measured by district wealth or by the principal's assessment of community social composition. MEAP reading and mathematics scores were found to be closely correlated with science scores. T-test results highlighted differences between schools scoring low and high on the test. Variables found to differ significantly between the two groups included years of science required, curriculum balance (attention to the various disciplines of science), curriculum appropriateness (a science curriculum tailored for college and non-college bound students), science spending per student, amount of science equipment, and the use of inquiry teaching techniques. Discriminant analysis revealed that three variables most successfully discriminated between low- and high-scoring schools: years of science required, academic preparation of teachers, and curriculum balance. Analysis of residuals saved from the regression of a measure of district wealth versus MEAP science scores identified variables which influence science achievement in absence of socioeconomic effect. Years of science required and curriculum appropriateness explained the most variance in the residuals (about 26%).