The goal of this research is to develop a model of teacher knowledge as a modifiable characteristic that mediates the effects of other teacher attributes (such as demographics and aptitudes), is modifiable due to interventions such as teacher professional development and teacher education, and has potential impacts on student and teacher outcomes in math and science. The methods of quantitative research synthesis will be used in this effort. Becker, Kennedy and O'Reilly, along with their staff and consultants Jakubowski and Southerland, will conduct a multi-part synthesis involving 1) studies of relationships among teacher characteristics, with a focus on teacher knowledge, 2) studies of teacher professional development and teacher education, and 3) studies relating teacher qualifications and knowledge to outcomes in the areas of math and science education.
Parts of the synthesis work are already underway thanks tobuild on two existing NSF-funded projects, headed by Becker and Kennedy (at Florida State University and Michigan State University) and O'Reilly (Abt Associates). The researchers will add to these projects a new synthesis of studies concerning the prediction and modification of teacher knowledge, with attention to the roles of both interventions and other teacher characteristics. This part of the project will include studies that examine teacher learning outside the context of professional development interventions. Such learning will include learning in teacher education programs (teacher preservice education) and the like.
These literatures will be merged by building on existing databases developed as a part of the two ongoing projects, using the tools of linked meta-analysis. Linking the findings of these projects will enable the researchers to build new models of teacher learning across the professional development continuum, from pre-service through in-service learning. The intellectual contribution of this research also will include the specification of gaps in the knowledge base concerning teacher learning, which will be identified by examining research on each link in the process of teacher knowledge development (as well as other factors that may impinge upon that process).
The empirical models will be constructed by way of multiple linked meta-analyses, also known as model-driven meta-analysis. Becker has developed methods for model-driven meta-analysis and also conducted model-driven meta-analyses on the prediction of child outcomes in divorcing families, sport performance outcomes, and the prediction of gender differences in science achievement. Eventually, the knowledge garnered from this synthesis can be tested in the context of teacher learning in professional development interventions, though that work is not a part of this specific proposal.