Key competencies acquired from STEM education: gender-differentiated parental expectations

The study investigates parents’ expectations regarding the key competencies children should acquire from STEM education and whether these expectations differ by the child’s gender. With globalization and the push for innovation, STEM education integrates science, mathematics, engineering, and technology to develop innovative thinking and real-world problem-solving. Parents strongly influence children’s STEM attitudes, motivation, and performance, especially in early years, and their expectations can be more impactful than direct involvement. Prior work suggests gender-biased expectations (e.g., boys seen as more suited for STEM; girls for language), but there is little consensus among parents on which STEM-related competencies matter most and limited research examining gender differences in such expectations. The study addresses two questions: (1) Which key competencies do parents most expect STEM education to cultivate? (2) Do parents’ expectations differ for boys versus girls?

Parents’ expectations are key predictors of children’s academic outcomes and STEM participation, sometimes surpassing other forms of involvement. Cross-cultural differences exist: Anglo-American parents emphasize conformity to external authority, whereas Mexican-American and many Western parents emphasize autonomy; Asian parents often stress academic achievement, social adaptability, and moral development. However, due to cultural specificity and measurement challenges, few studies directly examine parents’ expectations about children’s key competencies, and even fewer analyze gender differences. STEM-related key competencies, often aligned with 21st-century skills frameworks (e.g., P21 4C skills), include critical thinking, problem solving, creativity, communication, collaboration, information literacy, ICT skills, and scientific imagination. Asian frameworks (Japan, South Korea, China) similarly stress problem solving, creative thinking, self-management, communication, and innovation. Some work notes gaps between 21st-century frameworks and workplace-demanded STEM skills (e.g., engineering, systematic thinking, resource management), though these analyses are typically enterprise-oriented and focused on higher education rather than K-12. Literature on gender in STEM reports disparities in participation, fields of study, and attitudes, with mixed findings on performance differences; stereotypes and perceived abilities contribute to gaps. Parents often hold gendered expectations consistent with societal stereotypes, potentially influencing children’s STEM trajectories.

Design: Mixed-methods with a two-section survey and follow-up semi-structured interviews. Participants: 736 parents in China with children aged 3–12 enrolled in STEM courses (44% fathers, 56% mothers; children: 50.4% boys, 49.6% girls). Parents’ ages: 25–46; varied occupations (e.g., teachers, engineers, civil servants). Instruments: Parents’ Expectation Survey (PES) with demographics (parent and child) and ratings of 10 STEM-related competencies on a 7-point Likert scale (1=least important to 7=most important). The competencies (e.g., inquiring, creativity, problem solving, knowledge acquisition, expressing, thinking, anti-frustration, hands-on, concentration, cooperation) were compiled by STEM education experts referencing international frameworks. Face validity assessed by three STEM education experts; internal reliability Cronbach’s alpha=0.872. Procedure: Snowball sampling via a STEM education center in South China. After the survey, 46 parents who provided contact information participated in one-to-one interviews exploring definitions, reasons for expectations, and gender differences. Interviews were coded independently with high inter-coder reliability (Kappa=0.894) and consensus reached on themes. Data Analysis: Quantitative analyses in SPSS 26. Descriptive statistics, Exploratory Factor Analysis (principal component analysis; Quartimax with Kaiser Normalization). Suitability checks: KMO=0.846; skewness (−0.783 to −0.627) and kurtosis (0.341 to 1.582) within acceptable ranges. Factor loadings ranged ~0.686–0.891. Reliability for factors ranged approximately 0.785–0.878; overall alpha around 0.878. For gender differences, one-way ANOVAs compared expectations for boys vs. girls (by competency) and fathers vs. mothers; MANOVA tested interaction between parent gender and child gender on expectations. Qualitative data triangulated quantitative results and explored reasons behind patterns.

• Parents rated all 10 competencies highly (means ~5.96–6.26 on a 7-point scale). Inquiring had the highest mean; creativity had the smallest SD (greatest consensus).
• EFA identified four orthogonal factors explaining 80.564% of variance:
  1. Innovative factor: inquiring, creativity
  2. Social factor: cooperation, expressing
  3. Making factor: hands-on, anti-frustration, problem solving
  4. Learning factor: thinking, knowledge acquisition, concentration Factor loadings ~0.686–0.891; reliability acceptable (factor αs approx. 0.785–0.878; overall α≈0.878; initial instrument α=0.872). Parents emphasized factors in order: Innovative > Making > Learning > Social.
• Gender differences (child gender) in expectations (ANOVAs): • Expressing: girls > boys, F(1,735)=8.513, p=0.004, Cohen’s d=0.219 • Hands-on: boys > girls, F(1,735)=29.388, p<0.001, d=0.401 • Thinking: girls > boys, F(1,735)=8.563, p=0.004, d=0.217 • Learning competency: girls > boys, F(1,735)=14.400, p<0.001, d=0.282 • Concentration: girls > boys, F(1,735)=35.88, p<0.001, d=0.438 • No significant differences: inquiring, creativity, cooperation, anti-frustration, problem solving.
• Parent gender differences: Mothers expected higher anti-frustration competency than fathers, F(1,735)=5.698, p=0.017, d=0.169.
• Interaction (parent×child gender): significant only for expressing competency (p=0.025, η²=0.007). Mothers held similar expectations for boys and girls; fathers held higher expectations for girls in expressing.

Findings address both research questions. Parents’ expectations coalesce into four STEM-related competency factors and prioritize innovative competencies (inquiring, creativity), aligning with but also refining 21st-century frameworks by highlighting hands-on making and resilience (anti-frustration). Parents’ expectations are more focused and concrete than state-level frameworks, emphasizing practical expression, cooperation, and making rather than broader civic or cultural competencies. The Making factor is distinctive, underscoring STEM’s hands-on, learning-by-doing nature and valuing resilience when facing setbacks. Age-related considerations emerged (e.g., concentration for ages 3–12), and cultural context shaped expectations. Gender-differentiated expectations mirror common stereotypes: parents expect girls to excel in expressing, thinking, learning, and concentration, while boys are expected to excel in hands-on making. Fathers tend to prioritize external, utilitarian skills (girls’ expression; boys’ hands-on), whereas mothers emphasize internal capacities and resilience (anti-frustration). These patterns may reflect developmental perceptions (girls maturing earlier) and sociocultural norms, and they have implications for perpetuating gendered pathways in STEM.

This study identifies a parent-derived framework of STEM-related key competencies with four factors (Innovative, Social, Making, Learning) and documents systematic gender-differentiated expectations. It contributes a parent-centered perspective to competency frameworks, highlighting competencies often underrepresented in official models (e.g., hands-on making, anti-frustration) and age-salient needs (concentration). The results suggest the need to align STEM programs and evaluations with competencies parents value while addressing gender stereotypes that may shape children’s opportunities and self-concepts. Future research should survey both parents per child, compare parental expectations with teachers’ and children’s self-expectations, and examine cross-country differences to understand cultural influences and generalizability.

Only one parent per child was surveyed; dyadic (both parents) data were not collected. Comparisons with teachers’ expectations or children’s own expectations were not included. Results are from a single country (China), limiting cross-cultural generalizability; future cross-national comparisons are warranted.