This proposal focuses on science instruction at the elementary level where few research studies exist about effective models for teaching science. The goal of this study is to determine the degree to which there are meaningful differences in the quantity and quality of science instruction provided by a generalist, specialist, or blended model for classroom instruction; whether differences in instruction are associated with student outcomes; and which model is the most cost-effective based on financial and human resources. The study will target 80 teachers of fourth and fifth grade students in 42 schools in the Boston Public School District. The guiding principles detailed in the Massachusetts Science and Technology Engineering Curriculum Framework will serve as the operating definition of high-quality science instruction and the theoretical framework to guide classroom observations.
The four research questions that will guide this study include: 1. What are the characteristics of schools that do and do not employ science specialists? What are the characteristics of science specialists and classroom generalists? 2. How does the quality of science instruction provided by a science specialist, a classroom generalist, and a combination of specialist/generalist differ? How does the quantity of science instruction provided by a science specialist, a classroom generalist, and a combination of specialist/generalist differ? 3. If differences in instruction exist between the models, to what extent and in what ways are they associated with student learning? 4. What are the financial and human-resource costs associated with each model?
A mixed methods approach will be used to collect data through surveys, three face-to face interviews and three classroom observations. Data will be collected from both principals and teachers over a 2.5 year period. The survey instruments and protocols were developed by Horizon Research, Inc., and will be modified to meet the specific needs of this study. Fourth and fifth grade student test scores from the Massachusetts Comprehensive Assessment System, the state's assessment instrument, and kit-based tests will be used to determine the impact of the different models on student achievement. The study includes a well thought-out approach for analyzing data using an array of quantitative and qualitative methods (e.g. general descriptive statistics, hierarchical linear modeling, logistic regression models, one-way MANOVA estimations, coding and coding schemes, and atlas.it).
Broader Impacts The boarder impacts of this study reside in the significance and importance of results to district policymakers. While less that 30% of elementary students in the United States receive science instruction from specialists or a science specialist in addition to their regular classroom teachers, few studies exist that describe the effect of this approach on science teaching and student achievement. The fundamental issue that might be resolved through this study include which of three models (generalist, specialist, or blended) is the better one for teaching and learning science and at what cost. This information will help policymakers make better decisions about science instruction at the elementary level. Ways to disseminate lessons-learned to key education stakeholders will be distributed through multiple channels and pathways relevant to policymakers. Thus, this project adds to the limited research in this area, advances knowledge about science instruction, highlights the impact on student achievement, sheds light on cost effectiveness, and provides a tool for measuring science instruction at the elementary level.
Elementary science is important. Evidence indicates that students who do not have a solid exposure to science in the early years rarely make learning gains equal to those that did when they reach the secondary levels of schooling. Unfortunately, little research has been done that compares the effectiveness of the two models. As a consequence, school and district leaders are investing scarce resources in a strategy about which little is known and upon which much depends. This study examined science programs, instruction, and student outcomes at 30 elementary schools in a large, urban district in the northeast United States in an effort to understand whether there were meaningful differences in the quality, quantity and cost of science education when provided by a science specialist or a classroom teacher. We used student performance on the state’s mandated science achievement test as one measure of student learning, and students’ engagement in science lessons as another. We chose student engagement as an outcome because it was observable by researchers in the classroom, and although it is not a direct measure of student learning, it is a necessary prerequisite. A conceptual framework of the elementary science experience guided the study, and data was collected on all components of schools’ science instruction and science programs, including their costs, through interviews, observations, surveys, and school and district records. A simple comparison of the science specialist and classroom teacher models suggests that the science programs in schools where classroom teachers taught science were of higher quality overall and the difference was statistically significant; however there were no meaningful differences in the quality of the instruction that classroom teachers and science specialists provided. That said, among these 30 schools and across the two science program models, there were significant differences in the level of student engagement in their science lessons, and although some teachers provided better science instruction than others—and instructional quality accounted for 43% of the variance in students’ engagement in science lessons—the difference in quality could not be explained by a teacher’s classroom teacher or science specialist status, but rather to individual teachers within each group. Similarly, program costs were not consistently associated with student outcomes. Although poorly supported programs produced poor student outcomes, well-funded programs were not always associated with strong student outcomes, in fact sometimes the opposite was true – despite significant science program costs in schools where science specialists had a dedicated classroom, student outcomes were bleak. A comparison of the classroom teacher and science specialist models suggests that the classroom teacher model, while costing about 50% more than the science specialist model, achieves far better student outcomes. However, a comparison of the four most and equally effective classroom teacher and science specialist schools shows a much different outcome. These two science specialist schools delivered essentially the same outcome (.87 student engagement at a cost of $1,660) for 30% less than the two classroom teacher schools (.88 student engagement at a cost of $2,294). The science specialist schools’ higher program quality scores suggest that a strong, school-wide commitment to science, which is not easy to attain or monetize, may be needed in order for the science specialist model to be cost-effective. These data suggest, then, that there is no single answer to the question, who should teach elementary science, classroom teachers or science specialists? The answer is: it depends. Employing a specialist model guarantees that students will have some science instruction because lesson time is built into schools’ schedules, but they may not necessarily have high quality science instruction. And although there are heroic examples of specialists who are effective without a classroom, the data show that this piece of real estate has some benefit for the school, albeit at a cost. For example, specialists with a classroom take on a leadership role for science schoolwide where their colleagues without a classroom work longer hours just to perform their instructional duties. The same can be said of the teachers in these effective schools. Their principals made significant investments of both time and resources to provide them with a variety of supports, and these teachers delivered very positive student outcomes. One might ask if, in the absence of this level of commitment and support, and in light of competing pressures, would these teachers provide the same quality and quantity of science instruction? These findings provide valuable insights for those responsible for making the most cost-effective use of scarce resources in the service of students’ science learning. Regardless of the model, the data suggest a delicate balance between investing in science and valuing science. They are not the same thing, they are both critical, and whereas each one by itself can only do so much to advance students’ success in science, it appears that both together can accomplish more than one would expect.
