The contribution of attitudes toward school science in the explanation of achievement in STEM school subjects

Learning in the field of Science, Technology, Engineering, and Mathematics (STEM) is promoted in the educational policies of numerous countries. The achievement of students in Croatia in science and mathematics is low in comparison to their achievement in other school subjects (Burusic, Babarovic, Sakic, 2008). The debates about what determines achievement in STEM school subjects in primary school still persist. Several STEM-relevant variables show a significant association with achievement in math and science, including student level variables, school level variables and variables of child’s broader social environment (Hattie, 2009). The aim of this paper is to explain how school achievement in STEM school subjects among primary school students in Croatia is shaped by structurally different spheres of influence, namely by students’ home environment and family, leisure activities, hobbies and school activities. Huitt’s (2003) Transactional Model of the Teaching/Learning Process is used as a theoretical framework. The model proposes that the main variables that impact students’ academic achievement are home and other context variables, school level variables and classroom level variables. When student level characteristics are considered, a small gender differences in school achievement are found in mathematics (Frost, Hyde, & Fennema, 1994 ; Hyde, Fennema, & Lamon, 1990) and in science (Murphy & Whitelegg, 2006), with girls outperforming boys. Girls’ and boys’ also differ in their motivational orientation to STEM subjects, with girls prefering biological sciences and chemistry, whereas boys are more motivated in physical sciences (Weinburgh, 1995) and in computer-related tasks (e.g. Hayward et al., 2003). In predicting school achievement, it is important to consider not only the characteristics of the students, but also their social environment (Keith & Fine, 2005). When family background is considered, research has shown that socioeconomic status (including parental education, employment status and income) is the best predictor of school achievement (Dahl & Lochner, 2005 ; Milne & Plourde, 2006) and a good predictor of math and science achievement in primary school (Sirin, 2005). Likewise, among all other aspects of parents’ behavior, parents’ involvement in their children’s schooling is a powerful predictor of academic success (Gutman & Midgley, 2000 ; Fan & Chen, 2001). Furthermore, the impact of informal, out-of-school activities on STEM educational outcomes are also considerable (Braund & Reiss, 2004). Taking into account the importance of outside of school context and the family variables, in this research we are looking for evidence that achievement can also be improved by positive experience in formal school contexts. We assume that experience during the science and math school lessons can significantly add to explanation of achievement, above family influences and out of school activities. Therefore, the aim of this study is to identify the contribution of students’ attitudes to school science and experiences with STEM school subjects in explaining STEM school achievement, after controlling individual characteristics, family influences and experience in out-of-school activities. Methodology: Participants in the study were 360 primary school students attending grades 6 to 8 (age 12 to 15 ; M=13.32), among which 195 boys and 165 girls. Paper and pencil method was used to collect the data. Data was gathered on the students’ achievement in STEM school subjects, interest and attitudes in science in school and out of school and family related variables such as parental employment, parental education status and parental involvement in school. Data collection lasted 40 minutes and took place within regular classes in school. Scales used in the survey are mostly derived and adapted from the ASPIRES project (Archer, et al., 2013). Structural validity of all the used scales was checked and items that did not resemble expected and interpretive structure were removed to obtain clear factor structures. All scales have adequate reliability. We applied hierarchical regression analysis in order to predict achievement in STEM school subjects. The first block of predictors, named Family and Out- of- School STEM Attitudes and Interests consisted of student gender, school class grade, parental education status, parental ambitions/support (scale), parental attitudes to science (scale), positive images of scientists (scale), negative images of scientists (scale), and interest for science out of school (scale). In the second block, attitudes toward school science (scale) was added as a predictor to check if there is a significant increase explained variance in STEM school achievement. STEM school achievement is measured as a composite of school marks (GPA) in different STEM school subjects in the previous grade. Results: We performed a two-step hierarchical regression analysis to investigate the incremental validity of STEM school context in predicting STEM achievement, over and above the out-of-school STEM context. The regression model with eight out-of-school context variables entered in the first step and attitudes toward school science entered in the second step accounted for 22.5% of variance in achievement in STEM school subjects (R2=.23). The first and the second predictors’ blocks predicted students’ achievement in STEM school subjects to a statistically significant degree (F=8.10, p<0.05 ; F=11.33, p<0.05, respectively). By adding the attitudes toward school science in the second block the percentage of variance accounted for increased from 15.2% to 22.5%, (ΔR2=.073). It can be concluded that school context, or more specifically, attitudes toward school science, have an effect on achievement above and beyond the effects of out-of-school context. The significant predictors of achievement in STEM school subject were attitudes toward school science (β=.340), gender (β=.194), parents’ education (β=.194) and parental ambitions/support (β=.169). Engagement in science-related activities outside of school had no valuable explanation power in explaining STEM school achievement. The results are discussed in relation to prior research and theories in this domain.