A new study published in npj Science of Learning sheds light on the psychological factors influencing students’ decisions to pursue careers in science, technology, engineering, and mathematics. The research found that an individual’s interest in mathematics is a stronger predictor of STEM career choices than either their confidence in their mathematical abilities or their anxiety about the subject.
The study was motivated by the ongoing shortage of STEM professionals in the United Kingdom. Estimates suggest a shortfall of over 173,000 workers in the sector, which poses a challenge to economic growth and technological innovation. This issue is compounded by the underrepresentation of women in STEM fields, with only 27% of STEM workers being female, compared to 52% of the wider workforce.
While previous research has shown that maths achievement in adolescence predicts later participation in STEM, socio-cognitive and emotional factors, such as maths anxiety, self-efficacy, and interest, have also been linked to STEM engagement. However, few studies have examined how these factors interact to influence actual career choices, rather than just intentions to pursue STEM education. The researchers aimed to fill this gap by analyzing data from a large, longitudinal cohort.
“We were interested in this topic as we know the maths motivational attitudes and maths anxiety have an important but complex role in maths learning and achievement and we wanted to explore whether the same was true for STEM career and learning choices beyond compulsory education settings. This is particularly interesting in the context of there being a global and national (in the UK) STEM skills shortage,” said study author Rebecca Ferdinand, a PhD student at Royal Holloway, University of London.
The researchers used data from the Twins Early Development Study, a large-scale longitudinal study in England and Wales. The sample included 7,908 participants who provided self-reported data at multiple points: at age 16 (maths self-efficacy and interest), age 18 (maths anxiety), and age 21 (STEM career choice). STEM career choice was defined based on whether participants pursued a degree or apprenticeship in a STEM-related field. The study also accounted for prior maths achievement and socioeconomic status to examine whether these psychological factors independently influenced career decisions.
Participants completed standardized questionnaires assessing their confidence in solving mathematical problems (maths self-efficacy), their enjoyment of and engagement with maths (maths interest), and their feelings of nervousness and discomfort when engaging with maths-related tasks (maths anxiety). General anxiety levels were also measured to ensure that maths anxiety was assessed independently. Maths achievement was determined using General Certificate of Secondary Education (GCSE) exam scores, a standardized test taken at age 16 in the United Kingdom. Socioeconomic status was assessed based on parental education, occupation, and other demographic factors.
The results revealed that maths interest was the strongest predictor of STEM career choice. Participants with higher interest in maths were significantly more likely to choose a STEM career, even after accounting for maths achievement and socioeconomic status. Maths anxiety, on the other hand, was initially associated with lower odds of choosing a STEM career, but this relationship became non-significant when maths achievement was considered. This suggests that the negative impact of maths anxiety on STEM participation may be largely explained by its effect on students’ performance in mathematics, rather than a direct aversion to STEM fields.
Maths self-efficacy did not independently predict STEM career choice when analysed alongside maths interest and anxiety. While confidence in one’s maths abilities is often thought to play a role in academic and career decisions, this study suggests that interest may be a more decisive factor. This finding challenges assumptions from Social Cognitive Career Theory, which emphasizes self-efficacy as a key driver of career-related actions.
“We were surprised that maths self-efficacy didn’t predict STEM career choices alongside maths interest and anxiety,” Ferdinand told PsyPost.
However, given the strong correlation between self-efficacy and interest, it is possible that self-efficacy influences career choices indirectly by fostering greater engagement and enthusiasm for maths-related subjects.
Sex differences emerged in the patterns of influence for these psychological factors. Maths interest predicted STEM career choice for both males and females, reinforcing its importance across genders. However, maths anxiety was only a significant predictor for males, suggesting that male students who experience high levels of anxiety about maths may be particularly deterred from pursuing STEM careers.
One possible explanation is that males, who are often socialized to identify more strongly with maths, may experience a greater conflict between their anxiety and their self-concept, leading them to disengage from STEM paths. In contrast, while females reported higher levels of maths anxiety on average, this anxiety did not significantly influence their career choices.
Maths self-efficacy showed an interesting pattern: it was a significant predictor of STEM career choice for females in the unadjusted model, but this effect disappeared once maths achievement was taken into account. This suggests that, for females, confidence in mathematical ability may be tied more closely to actual performance in maths, whereas for males, interest in the subject may be a more dominant factor in career decisions. These findings align with previous research showing that self-perceptions of ability play a larger role in shaping academic choices for females, while interest is a stronger motivator for males.
“Our findings highlight that maths interest might play a more central role in STEM career choice than maths anxiety or self-efficacy,” Ferdinand said. “It also suggests that the gender gap in STEM could be at least partially due to a discrepancy in adolescent maths interest, rather than in maths anxiety or self-efficacy.”
While the study identifies associations between maths interest, anxiety, and career choices, it does not establish causation. Other factors, such as parental influence, teacher encouragement, or exposure to STEM role models, may also play a role in shaping students’ career trajectories. Further research is needed to explore how these external influences interact with psychological factors to determine STEM participation.
“We only looked at these factors in late adolescence (self-efficacy and interest at 16, maths anxiety at 18),” Ferdinand noted. “But early childhood experiences of maths anxiety could affect maths interest and self-efficacy levels in adolescence, creating more complex domino effect on STEM career choices. Other social, emotional, and motivational factors may also play a role, like parental or teacher support or financial goals.”
The long-term goal is to understand individual differences in education choices and learning, as this could be an important area of research for improving access, social mobility, and the economic climate.
The study, “Mathematics interest, self-efficacy, and anxiety predict STEM career choice in emerging adulthood,” was authored by Rebecca Ferdinand, Margherita Malanchini, and Kaili Rimfeld.
