School Level Factors Contributing to Reductions in Science Achievement Gaps

• In Event: Understanding Emerging Opportunity Structures in STEM Education: Research Efforts to Explore School-Level Innovations

Abstract

Disparities in science achievement have been widely reported for many years. However, this knowledge has not been sufficient to promote action. The remoteness of TIMSS data from the daily work of science educators has recently been jostled as an outgrowth of federal accountability mandates. With science testing required in elementary grades, the stipulation that these data be made publically available, and that these data are to be disaggregated by student demographics highlights the salience of disparities and inequities corresponding to ethnicity, ELL status, social class and so on – realities that are literally quite close to home. NAEP data has shown that Connecticut has had among the largest science achievement gaps; annual statewide reports highlight district and school level performance in ways that are difficult to ignore.

Much of the school effectiveness research has focused on language arts and mathematics achievement. Those studies treat the school as the level of analysis, simultaneously attending to student outcomes but examining how school level organization and leadership are associated with student performance. Because science is first tested in at grade five, this measure represents all the science a child has learned in the first ten or so years of life. While fifth grade teachers are certainly on the front line when test results are announced, a more systemic consideration of inputs is required. Attending to program and system concerns appeared in the original National Science Education Standards but with rare exceptions, science education reform has largely maintained an emphasis on curriculum and instruction. A collaborative research and development project between Connecticut and Florida is designed to identify the congruities (and contradictions) of school effectiveness research when that literature is applied to science achievement.

Multiple regression analysis provides a quantitative mechanism for identifying schools with varying levels of success in terms of equitable science outcomes. Among schools serving culturally and linguistically diverse populations, we are examining school level factors associated with lesser versus greater achievement disparities. Multilevel analyses of teachers’ perceptions of school climate, of neighborhood information from the census bureau and other government sources, of teacher reports of school/community relationships, along with leadership perspective will verify how accurately other school effectiveness findings apply to students science outcomes in urban and diverse schools. This information is complemented by interviews of school administrators, science teachers and community members. Theoretical frameworks informing this study’s design include social capital, organizational trust, funds of knowledge, distributed leadership, among others. Key dimensions include the deliberate attention to closing achievement gaps, the emphasis on a district-wide scope of investigation, and the possibility that school organization and leadership contains specific characteristics distinct to science. By virtue of this five-year study, the project is poised to examine how shifts in leadership (through personnel changes or policy adjustments) influence science test scores. Ultimately, this in situ examination of 150 schools will provide insights that can inform interventions for other settings.

Authors

• John Settlage, University of Connecticut

• Malcolm B. Butler, University of Central Florida