Circumscription and Compromise theory suggests self-concept and sex stereotype explain occupational preferences, including preferences for science, technology, engineering and mathematics (STEM). Support exists for sex differences between males and females in both science degrees and science careers. The main thrust of observed sex differences in science lies in the development of occupational interest, as it has been suggested females are encouraged away from science due to stereotypes and social pressure. The present study evaluates high school juniors and seniors (n = 295) to explore their preference for science as indicated by science motivation, attitude, academic experience, and interest. Latent Profile Analysis was used to model profiles of preferences for science with a person-centered approach. Then, the impact of self-concept variables was explored and four profiles of science interest were identified. Sex differences were identified based on science interest, but were not always in favor of males. Covariate analysis indicates vocabulary ability and personality as significantly different for students in the high science interest profile. Implications of these results and future research directions are discussed.

Data-driven decision making is an integral part of higher education and it needs to be rooted in strong methodological and statistical practices. Key practices include the use and interpretation of effect sizes as well as a correct understanding of null hypothesis significance testing (NHST). Therefore, effect size reporting and interpreting practices in higher education journal articles represent an important area of inquiry. This study examined effect size reporting and interpretation practices of published quantitative studies in three core higher education journals: Journal of Higher Education, Review of Higher Education, and Research in Higher Education. The review covered a three-year publication period between 2013 and 2015. Over the three-year span, a total of 249 articles were published by the three journals. The number of articles published across the three years did not vary appreciably. The majority of studies employed quantitative methods (71.1%), about a quarter of them used qualitative methods (25.7%), and the remaining 3.2% used mixed methods. Seventy-three studies were removed from further analysis because they did not feature any quantitative analyses. The remaining 176 quantitative articles represented the sample pool. Overall, 52.8% of the 176 studies in the final analysis reported effect size measures as part of their major …