Larry J. Grabau

Science engagement and science achievement in the context of science instruction: a multilevel analysis of U.S. students and schools

ABSTRACT Using data from the 2006 Program for International Student Assessment (PISA), we explored nine aspects of science engagement (science self-efficacy, science self-concept, enjoyment of science, general interest in learning science, instrumental motivation for science, future-oriented science motivation, general value of science, personal value of science, and science-related activities) as outcomes and predictors of science achievement. Based on results from multilevel modelling with 4456 students nested within 132 schools, we found that all aspects of science engagement were statistically significantly and positively related to science achievement, and nearly all showed medium or large effect sizes. Each aspect was positively associated with one of the (four) practices (strategies) of science teaching. Focus on applications or models was positively related to the most aspects of science engagement (science self-concept, enjoyment of science, instrumental motivation for science, general value of science, and personal value of science). Hands-on activities were positively related to additional aspects of science engagement (science self-efficacy and general interest in learning science) and also showed a positive relationship with science achievement.

Aspects of Science Engagement, Student Background, and School Characteristics: Impact on Science Achievement of U.S. Students

OF DISSERTATION ASPECTS OF SCIENCE ENGAGEMENT, STUDENT BACKGROUND, AND SCHOOL CHARACTERISTICS: IMPACTS ON SCIENCE ACHIEVEMENT OF U.S. STUDENTS Science achievement of U.S. students has lagged significantly behind other nations; educational reformers have suggested science engagement may enhance this critical measure. The 2006 Program for International Student Assessment (PISA) was science-focused and measured science achievement along with nine aspects of science engagement: science self-efficacy, science self-concept, enjoyment of science, general interest in learning science, instrumental motivation for science, future-oriented science motivation, general value of science, personal value of science, and science-related activities. I used multilevel modeling techniques to address both aspects of science engagement and science achievement as outcome variables in the context of student background and school characteristics. Treating aspects of science engagement as outcome variables provided tests for approaches for their enhancement; meanwhile, treating science achievement as the outcome variable provided tests for the influence of the aspects of science engagement on science achievement under appropriate controls. When aspects of science engagement were treated as outcome variables, gender and father’s SES had frequent (significant) influences, as did science teaching strategies which focused on applications or models and hands-on activities over-and-above influences of student background and other school characteristics. When science achievement was treated as the outcome variable, each aspect of science engagement was significant, and eight had medium or large effect sizes (future-oriented science motivation was the exception). The science teaching strategy which involved hands-on activities frequently enhanced science achievement over-and-above influences of student background and other school characteristics. Policy recommendations for U.S. science educators included enhancing eight aspects of science engagement and implementing two specific science teaching strategies (focus on applications or models and hands-on activities). Focused implementation of these research findings could enhance both science engagement and science achievement of U.S. students. I identified five key limitations of my research project: the age of the dataset, the lack of racial/ethnic identifiers, the low proportion of student-level variance accounted for by multilevel models with aspects of science engagement as outcome variables, the lack of class-level measures, and the lack of inclusion of students’ epistemological and fixed/flexible beliefs. These limitations provide opportunities for further investigations into these critical issues in science education.