Abstract | Keywords | Export | Availability | Bookmark

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

STEM (“Science, Technology, Engineering and Mathematics,”) is increasingly becoming a hot topic due to the dynamic and evolving nature of new sciences and technologies. According to the STEM pipeline analogy, women tend to "leak" out of the pipeline quicker and earlier than men which results in gender divides at various stages of this pipeline. In the context of Rwanda, while the Rwandan government has passed various policies to achieve gender parity, there still remains an under-representation of women in STEM careers and higher education. This study firstly identified a few theoretical and key determinants of motivation to pursue STEM in higher education. These indicators are defined as Self-efficacy in school disciplines, Spatial Ability, Gender Stereotypes, and Responsibilities and Constraints. The methodology to investigate the effect of these determinants on motivation to pursue STEM involved conceptualising a logistic MGSEM (Multi-group Structural Equations Modelling). To operationalise this, a pre-collected survey dataset by the Centre for Sociological Research at KU Leuven, VVOB – Education for development as well as the Rwandan Association for Women in Science and Engineering (RAWISE), was used. The survey took place in 20 secondary schools in Kayonza, Rwanda, selected based on a stratified random sample, and produced 915 respondents from grades S1, S3 and S6. A latent variable structure was constructed for 3 of the 4 determinants (Responsibilities and Constraints was sum-scaled instead, since all observed variables within it were binary). Next, a logistic regression on the binary outcome of pursuing STEM in higher education, was estimated using the determinants as predictor variables. And finally, the final model was compared across male and female students in a logistic MGSEM. This was all possible via the R package for complex SEM analyses: lavaan. No statistically significant differences were found between male and female students based on the regression results of this analysis, which highlights the importance of further modelling techniques or other strategies to measure the behaviours of female students when it comes to STEM. A simplified marginal effects analysis was conducted to find that female and male students both have a higher probability of pursuing STEM in higher education if they possess higher self-efficacy. However, for female students, this probability decreases if they held gender biases towards STEM activities (boys are better at STEM than girls). Overall, this study can help facilitate research on gender differences in STEM fields in Rwanda, and provide a few valuable insights into the behaviours of secondary to high school students when it comes to the decision-making of their future studies and careers. This study also uses relevant and powerful sets of modeling tools, like a logistic SEM to measure gendered differences in behaviours, which can be built on and adapted by future researchers and analysts.