Individual and contextual effects on science identity among American ninth-grade students (HSLS:09): hierarchical linear modeling

ABSTRACT Background Research has shown that students in the United States have been developing negative perceptions of science as a field of study, which is associated with low science identity. The science identity can be influenced by a variety of personal demographic predictors, non-cognitive science-related variables (e.g. science self-efficacy and science interest), and contextual predictors (e.g. school locale and geographical location). Purpose The current study explored the variability in the effect of student-level factors (i.e. gender, SES, ethnicity, science self-efficacy, science utility, and science interest) and school-level factors (i.e. school type, locale, and geographical location) on ninth-grade students’ science identity. Sample Data on 16,216 ninth-grade students in the United States were obtained from a large-scale national database, the High School Longitudinal Study of 2009. Design and Methods Three models were tested using Hierarchical Linear Modeling. Results The findings showed that student-level factors accounted for a significant amount (36%) of the variability in students’ science identity and a much smaller amount (< 5%) of the variability in identity was due to the variation between schools. Students’ science identity was bolstered by a higher level of science self-efficacy, utility, interest, and socioeconomic status. Females, African Americans, and Hispanics had lower levels of science identity. Science identity was lower among students in rural schools and town schools compared with students in city schools and higher among students in private schools compared with students in public schools. Students in the Northeast and South had higher science identity compared to students in the West. Conclusion Establishing valuable contributions in the STEM field is strengthened when students hold constructive science identities. The current study found that various individual and contextual variables formulate students’ science identities. These influences are salient in rural and town public schools with poor SES status.