The research goal of this project is to evaluate whether an early childhood science education program, Head Start on Science, implemented in low-income preschool settings (Head Start) produces measurable impacts for children, teachers, and parents. The study is being conducted in eight Head Start programs in Michigan, involving 72 classrooms, 144 teachers, and 576 students and their parents. Partners include Michigan State University, Grand Valley State University, and the 8 Head Start programs. Southwest Counseling Solutions is the external evaluator.

The study is determining the efficacy of the Head Start on Science curriculum in two models, one in which 72 teachers participate in professional development activities (the intervention), and another in which 72 teachers receive the curriculum and teachers' guide but no professional development (the control). The teacher study is a multi-site cluster randomized trial (MSCRT) with the classroom being the unit of randomization. Four time points over two years permit analysis through multilevel latent growth curve models. For teachers, measurement instruments include Attitudes Toward Science (ATS survey), the Head Start on Science Observation Protocol, the Preschool Classroom Science Materials/Equipment Checklist, the Preschool Science Classroom Activities Checklist, and the Classroom Assessment Scoring System (CLASS). For students, measures include the "mouse house problem," Knowledge of Biological Properties, the physics of falling objects, the Peabody Picture Vocabulary Test-Fourth Edition, the Expressive Vocabulary Test-2, the Test of Early Mathematics Ability-3, Social Skills Improvement System-Rating Scales, and the Emotion Regulation Checklist. Measures for parents include the Attitudes Toward Science survey, and the Community and Home Activities Related to Science and Technology for Preschool Children (CHARTS/PS). There are Spanish versions of many of these instruments which can be used as needed. The external evaluation is monitoring the project progress toward its objectives and the processes of the research study.

This project meets a critical need for early childhood science education. Research has shown that very young children can achieve significant learning in science. The curriculum Head Start on Science has been carefully designed for 3-5 year old children and is one of only a few science programs for this audience with a national reach. This study intends to provide a sound basis for early childhood science education by demonstrating the efficacy of this important curriculum in the context of a professional development model for teachers.

This exploratory study by researchers at WestEd is examining ways that teacher-level factors (including teacher background variables and instructional practices) and student-level factors (such as self-rated mathematics interest and proficiency), and interactions among these factors, are associated with American Indian/Alaska native (AI/AN) student academic achievement in middle grades mathematics. The ultimate goal is to identify malleable factors that, if changed, could improve teachers' practices and AI/AN student achievement in mathematics.

The study has two main phases. Phase I is exploring data from the 2007 National Indian Education Study (NIES), which compiled information on student performance on NAEP assessments and collected information from a large sample of AI/AN students and their teachers through individual questionnaires. The NIES information links NAEP performance data and NIES survey data in ways that will assist understanding of how particular teacher-level factors and student-level factors (and the interactions between these two factors) relate to student learning. Hierarchical linear modeling (HLM) is being used to study the relationship between student/teacher-level factors and student performance. Phase II of the study involves qualitative analysis of data from several hundred interviews and classroom observations in selected schools at three sites in Alaska, Arizona, and New Mexico to enable deeper understanding of AI/AN contexts of teaching and learning and why particular teacher-level and student-level factors are/are not associated with student performance on the NAEP mathematics assessment as analyzed in Phase I.

Study data sources and application of multi-level modeling techniques will shed light on how teachers in particular cultural settings apply standards, adapt and implement curriculum, and assess their students in ways that promote student learning and achievement. The long-term payoff from this work will be enhanced understanding of ways to provide culturally responsive STEM education and increased performance and participation of AI/AN students in STEM careers and STEM-affected social and personal decision-making.