Perceived employability and career readiness among STEM students: does gender matter?

ABSTRACT

Gender equity is an area of concern within STEM, where women are underrepresented in education and career roles. Increasing the representation of women in STEM and removing negative gender stereotypes are necessary actions, both of which can be fostered during formal education through employability initiatives and in the workplace. This study focused on the university setting and explores the perceived employability beliefs of students in STEM (including medicine), particularly their career identity and commitment, learning mindset, awareness of career and how their learning relates to their future career. The 12,708 student responses to an online questionnaire illustrate significant gendered differences in employability beliefs. Female students in STEM have significantly higher confidence in career identity and commitment, and relevance of learning compared to males, although they feel less confident in their reconsideration of career commitment and occupational mobility. In comparison to non-STEM students, women in STEM report higher confidence in career identity and commitment, reconsideration of career commitment, willingness and ability to learn, relevance of learning, career exploration and awareness and occupational mobility. Employability interventions are needed within the STEM educational context to promote long-term equal opportunities for all genders in the workplace.

KEYWORDS:

• STEM
• gender
• career
• employability
• confidence

Introduction

Promoting gender equity in education and the labour market is critical for any economy. While gender equity has been achieved in some areas, it is still of concern within STEM (Science, Technology, Engineering and Mathematics), where women represent less than 30% of the total population in STEM fields of study and careers including countries like America and Australia (Australian Academy of Science, 2019; Dockery & Bawa, 2018; Wang & Degol, 2017). Researchers from different countries use various definitions of these fields of study such as STEM (science, technology, engineering and mathematics), STEMM (which includes medicine) (see for example Bennett et al., 2020a), and STEAM (with arts) (see for instance Dockery et al., 2021). In this paper, we adopt the 2015 Science in Australia Gender Equity definition of STEM, which includes Science, Technology, Engineering, Mathematics and Medicine. As the Office of the Chief Scientist (2020) states, health qualifications rely on scientific knowledge and practices and hence are included in the broader definition of STEM.

Besides the underrepresentation of women, there is also the concept of a leaky pipeline in STEM (Metcalf, 2010) whereby the number of young women keeps on shedding as they move from school to university and further when it is time to choose their careers. While various policies such as the ‘women in STEM decadal plan’ (Australian Academy of Science, 2019) have been instigated to correct the underrepresentation of women in STEM by encouraging more women into the field, there is also emerging evidence that women in STEM face labour market inequities in terms of gender pay gaps, higher unemployment, and fewer opportunities for career progression (Australian Academy of Science, 2019; Dockery & Bawa, 2018; Office of the Chief Scientist, 2020; Professionals Australia Survey, 2018). Women also encounter additional barriers to entry, retention and career progression than men in STEM (Australian Academy of Science, 2020) and are more likely to leave their STEM careers (Holman et al., 2018). Thus, ‘STEM organisations should aim to improve gender equality at work both numerically (improving women’s representation) and normatively (removing negative gender stereotypes)’ (van Veelen et al., 2019, p. 15).

Unfortunately, gender equity is expected to worsen due to the recent global pandemic. The Australian Academy of Science (2020) notes that women are already in the minority in STEM professions and may be at higher risk of job insecurity leading to negative career outcomes. It is crucial that universities understand not only the characteristics but also the employability beliefs of STEM students to help foster positive career trajectories and decision-making practices, particularly for women. This paper investigates whether gender impacts on perceived employability and career readiness among STEM students. It is important to study what women think or feel about their employability skills and future career because self-beliefs can affect the learning and performance in their university studies and impact on career decisions (OECD, 2015).

STEM

STEM skills are critical to the growth of any economy and a broader representation of the population in these fields is required. Enhancing the access of underrepresented groups to STEM can help boost economic growth and prosperity and is particularly important in the post-COVID era (PCAST, 2020) because a ‘STEM-literate public will be better equipped to handle rapid technological change … ’ (PCAST, 2018, p. V;7) and ‘the innovation capacity … depends on an effective and inclusive STEM education ecosystem’ (PCAST, 2018, p. V;7).

Despite an increasing emphasis on STEM skills, there remains an underrepresentation of women in STEM fields of study and careers. The OECD (2017) states that ‘while young women in OECD countries now obtain more years of schooling than young men, on average, girls are much less likely to study in the lucrative science, technology, engineering and mathematics (STEM) fields’ (OECD, 2017, p. 3). Women are severely underrepresented in engineering and IT, while there is a general overrepresentation of women in health (Holman et al., 2018; Makarova et al., 2019; WEF, 2017). Hinting at biological difference, Wang and Degol (2017) document six reasons for underrepresentation of women in STEM, citing cognitive ability, relative cognitive strengths, occupational interests or preferences, lifestyle values or work-family balance preferences, field-specific ability beliefs, and gender-related stereotypes and biases (also see Makarova et al., 2019). Gender stereotypes remain prevalent in many countries despite high gender equity (Miller et al., 2015).

Gender stereotypes provide generalisations about individuals based on their gender. These are the beliefs and attitudes towards behaviours of men and women and their roles in families and careers (Makarova et al., 2019; Renfrow & Howard, 2013). Gender stereotypes from early influencers including parents and teachers can inhibit girls’ performance and adversely affect their career choices (Ertl et al., 2017). Prior to a STEM career, the educational context contributes significantly to shaping future graduates’ attitudes, skills and behaviours and thus requires consideration.

There are many educational influences that affect career choices for young women (Bennett et al., 2022). Wang (2012) states how classroom experiences of students can predict their academic choices and career aspirations while Kaleva et al. (2019) argue that differences in self-efficacy beliefs influence career choices. Negative stereotypes can emerge through classroom assessments and male-dominated classrooms and these stereotypes can affect women despite having good grades (Ertl et al., 2017). Cundiff et al. (2013) found stronger gender stereotypes in science were linked to lower identification and career aspirations among women, while also linked to stronger identification and higher career aspirations among men. O’Dea et al. (2018) suggest that a possible reason girls do not choose STEM careers, despite having grades equivalent to those of their male peers when they graduate, could be increased male competition in STEM careers than non-STEM careers and the stereotype threats in these careers. Wang et al. (2013) present an alternative view, arguing that females with the same high maths abilities as males generally have higher verbal abilities than their male counterparts. As a result, females may explore both STEM and non-STEM careers and pursue a broader choice of careers than males.

Even women who pursue STEM careers can be affected by these threats (Ertl et al., 2017). Gender stereotypes relate to gender inequality in achievements and such beliefs further affect the interests and aspirations for men and women to enrol in STEM subjects (Makarova et al., 2019; Nosak et al., 2009). This can have a negative impact on women’s academic performance in the short-term and loss of identity in these fields in the long-term (Jones et al., 2013). Lack of confidence among women in technical fields has been aligned with women’s lack of success in these fields (Catherine et al., 2010; Ypulse survey, 2018) and is exacerbated by the lack of women role models in STEM fields of study and careers. The underrepresentation of women in STEM leads to further spread of the gender stereotypes affecting the career interests and choices of female students (Makarova et al., 2019). To address the underrepresentation of women in these fields and careers there is a need to systemically remove barriers at each level of the STEM pipeline (Australian Academy of Science, 2019) including at the macro contextual level, organisational and individual level (Bolzani et al., 2021; Herman, 2015).

Employability within the context of higher education ‘relates to the process by which we prepare students to negotiate graduate life and work’ (Bennett, 2019, p. 34). As Herman (2015) notes, employability is relevant throughout the career lifespan. As an underachievement of women in STEM can have consequences for their participation in the labour market and economic growth (OECD, 2015), it is important to reach gender equity in this area. In this study, we analysed the field-specific beliefs of young women in STEM to identify gender differences among higher education students’ perceived employability.

Theoretical framework

The study was grounded in Social Cognitive Careers Theory (SCCT) (Lent et al., 1994), which derives from Bandura’s (1996) Social Cognitive Theory. SCCT describes that people generate career interests by developing confidence in activities related to their interests, with outcome expectations validated through expended effort (Betz & Hackett, 1983; Johnson & Muse, 2017). In the pre-career phase, confidence relates to both work and study (e.g., both academic self-efficacy and career orientation). Santos (2020) argues that there is no right or wrong combination of skills, attitudes or attributes; rather, students need to make decisions and express their employability thinking within a metacognitive frame.

Advocating the importance of perceived employability for both decision making and student performance, Baruch and Peiperl (2000) highlight the development of self-awareness, self-esteem, self-efficacy and confidence as inner-value capitals. In line with SCCT, Donald et al. (2019, p. 599) take a broader view of capitals, highlighting the importance of human capital to undergraduates’ perceived employability and encapsulating ‘social capital, cultural capital, psychological capital, scholastic capital, market-value capital, and skills’. The study reported here maintained the broader human capitals approach and considered students’ confidence in relation to self- and career-awareness: perceived study relevance and learning mindset, career identity, career commitment and career awareness. The five constructs are summarised to follow, with items listed in Appendix 2.

Students’ career identity and commitment were measured using an 8-item scale developed by Mancini et al. (2015). The first four items measure the extent to which students identify with their study area and the remaining items measure the extent to which students would change their study area if they could. Items employ a 5-point Likert scale.

Students self-assessed their willingness and ability to learn – their cognitive openness to maintaining their knowledge, skills and abilities – using Coetzee’s (2014) 7-item continuous learning orientation scale, which employs a 6-point Likert scale.

Perceived program relevance refers to students’ confidence that they can recognise the relevance of their learning tasks and integrate theory and practice into workplace settings. Perceptions of programme relevance influence student motivation (Kember et al., 2008), study retention and completion (DeLottell et al., 2010) and knowledge retention (Malau-Aduli et al., 2013). Students responded to a 4-item scale adapted from Smith et al. (2014), which employs a 5-point Likert scale.

Career exploration & awareness and career mobility were measured using an 8-item scale developed by Lent et al. (2016) and focused on decisional self-efficacy and decisional coping efficacy. Items were measured using a 10-point Likert scale.

Instruments and approach

Ethical approvals were obtained before commencement (approval HRE2017-0125). Students received an information sheet and assurance of anonymity, and they completed a consent form. Students chose whether or not to include their online tool responses in the research dataset and this decision did not affect their access to the tool and associated resources. The study was conducted in two phases, described below.

Phase one

Phase One of the study employed Bennett’s validated online survey of perceived employability, which has been successfully applied within recent STEM research (Bennett et al., 2020a; Bennett et al., 2020b; Bennett et al., 2021). Grounded in SCCT, the instrument measures domains of career and study confidence as shown in Appendix 2; these are combined to create an indication of overall perceived employability. The reliability for each construct has been previously estimated, with all constructs having alphas over 0.70 (Bennett et al., 2022). The online survey takes ∼30 minutes to complete, with the initial outcome for students being a personalised profile report with further information and embedded links to developmental resources. A total of 13,110 survey responses were received. Once incomplete responses had been removed, 12,708 usable responses were retained and analysed. Data were inputted to SPSS v. 26.

Phase two: supplementary focus group

Once Phase One data analysis was complete, a student focus group was conducted to ensure that our interpretation of the survey responses was consistent with the student perspective (Jo et al., 2022; Zhang et al., 2022). The focus group involved two male STEM students and two female non-STEM students. The session lasted 90 minutes and was semi-structured.

One researcher led the session, and a second researcher was a silent observer. The interview prompts were printed cards on which items within the scale under discussion were reframed as statements. For example, identification with commitment included the statement ‘Thinking of myself as a professional in my discipline helps me to understand who I am’ (Mancini et al., 2015).

A multi-step approach inspired by Ananthram and Chan (2016) was employed to minimise social desirability bias and encourage truthful responses. In summary, participants volunteered; experienced interviewers conducted the focus group; the interview setting was familiar to participants; the time and place of the focus group was convenient for participants; there were no right or wrong answers; and participants were encouraged to use anecdotes and experiential evidence to support their views.

The interview transcript was analysed using thematic content analysis, which enabled the systematic, replicable compression of text into fewer content categories based on explicit rules of coding (Weber, 1990) and inspection of the data for recurrent instances (Wilkinson, 2011). The researchers first coded inductively in line with the themes from Phase 1, and then looked for any new themes.

Findings and discussion

The demographic profile of participants is summarised in Table 1. The sample comprised undergraduate students: 46 percent in Year 1, 17 percent in Year 2, 22 percent in Year 3, and 15 percent in Year 4. Consistent with other research within STEM, female students dominated the health field of study as well as the natural and physical sciences whereas the other STEM fields were dominated by male students. Within the non-STEM fields, males and females were equally represented within management and commerce whereas other non-STEM codes were dominated by female students (See for instance, WEF (2017), Makarova et al. (2019), Jones et al. (2013), Ertl et al. (2017), and Office of the Chief Scientist (2020)). According to the OECD (2017), these disparities are persistent.

Table 1. Demographic profile (n = 12,708) of participants from Phase One.

Field of study and code	Male	Female	All
Natural and Physical Sciences (code 1)	169	263	432
Information Technology (code 2)	1367	303	1670
Engineering and Related Technologies (code 3)	320	71	391
Architecture and Building (code 4)	364	219	583
Agriculture, Environmental and Related Studies (code 5)	30	28	58
Health (code 6)	705	2157	2862
STEM (1 + 2 + 3 + 6)	2,561	2,794	5355
Education (code 7)	164	734	898
Management and Commerce (code 8)	1686	1966	3652
Society and Culture (code 9)	436	905	1341
Creative Arts (code 10)	299	478	777
Food, Hospitality and Personal Services (code 11)	3	13	16
Mixed Fields (code 12)	8	20	28
Non-STEM (4 + 5 + 7 + 8 + 9 + 10 + 11 + 12)	2,990	4,363	7353
All	5551	7157	12,708

A demographic profile of the focus group participants is presented in Table 2.

Table 2. Demographic profile of participants in phase two (n = 4).

#	Gender	Field of study	Year of Study
STEM – 1	Male	Electrical engineering (code 3)	Year 4
STEM – 2	Male	Physics (code 1)	Year 3
Non-STEM 1	Female	Economics (code 8)	Year 3
Non-STEM 2	Female	Economics & Finance (code 8)	Year 3

Exploratory factor analysis, reliabilities and correlations

All 31 items were subject to exploratory factor analysis (EFA) and reliability assessments. EFA was performed to establish the underlying structure of the robustness of the assessed variables. Reliability assessments were used to assess the internal consistency of all the scales.

The EFA procedure employed the principal components method for extraction, and factors with eigenvalues greater than one were retained. Orthogonal rotations were performed with the Varimax option because of the technique’s success in obtaining orthogonal rotation of the factors for the purposes of regression and other prediction techniques (Hair et al., 1998). Given that the EFA sample size was 12,708, Hair et al. (1998) suggest that conservative factor loadings of greater than ± . 40 should be considered at the .05 level. All 31 items entered into the EFA were retained and the items are represented in Appendix 2.

The EFA identified that all factors were unidimensional. Reliability for each construct was estimated using Cronbach’s alpha coefficient (Cronbach, 1951) with a reliability measure of .70 considered acceptable (Nunnally, 1978). Shown in Table 3, all constructs had alphas over 0.80 indicating acceptable internal consistency and confidence in generalisability.

Table 3. Means, correlations and reliabilities.

	Mean (s.d.)	1	2	3	4	5	6
1. Identification with career commitment	3.872 (0.781)	0.837
2. Reconsideration of career commitment	2.602 (1.029)	−0.167 **	0.829
3. Understanding the relevance of learning	4.678 (0.794)	0.412 **	−0.120 **	0.883
4. Perceived programme relevance	4.100 (0.559)	0.409 **	.199 **	0.527 **	0.805
5. Career exploration and awareness	7.897 (1.422)	0.453 **	−0.200 **	0.584 **	0.521 **	0.943
6. Occupational mobility	7.379 (1.628)	0.294 **	−0.048 **	0.498 **	0.399 **	0.677 **	0.870

Notes: 1. Bold values on the diagonal are Cronbach alphas; 2. **p < 0.01.

Multivariate analysis to assess gender differences

In this paper, we focus on the six scales shown in Table 3. We report the means of the factor scores for these measures by gender within STEM and Non-STEM student cohorts. We follow the convention of referring to results as highly significant if the p-score is less than 0.01; moderately significant for a p-score from 0.01 to 0.05 and weakly significant for a p-score from 0.05 to 0.10 (as significant at the 1 percent with ***, 5 percent with ** or 10 percent with * level, respectively).

Table 4 presents the gender mean score differences in the confidence levels of the six assessed employability constructs in STEM. From Table 4, it can be seen that female students in STEM demonstrate significantly higher confidence in career identity and commitment and relevance of learning than their male counterparts. The higher levels of confidence imply that women in STEM identify with the STEM subjects and feel more committed to their careers than is suggested in traditional gender stereotypes. This is consistent with Jones et al. (2013), where women reported higher perceptions of their engineering abilities than men. Yet in the Jones et al. (2013) study, women gave lower ratings for their expectancy of future success; this contrasts with the present research.

Table 4. Gender mean score differences in the confidence levels in STEM (n = 5355).

Constructs	Female (n = 2794) Mean (s.d.)	Male (n = 2561) Mean (s.d.)	Mean difference	p value
1. Career Identity and commitment	3.987 (.738)	3.897 (.781)	.089***	0.000
2. Reconsideration of career commitment	2.485 (1.061)	2.653 (1.050)	-.167	0.000
3. Willingness and ability to learn	4.755 (.771)	4.729 (.820)	.026	0.222
4. Relevance of learning	4.219 (.514)	4.119 (.564)	.099***	0.000
5. Career exploration & awareness	8.019 (1.37)	7.992 (1.412)	.0269	0.479
6. Occupational mobility	7.341 (1.628)	7.568 (1.649)	-.227	0.000

Notes: The mean difference is (female mean score – male mean score). +ve value means female score > male score which indicates, on average, females are more confident in that literacy skill. P< =  0.1 = 10% los, p< =  0.05 = 5% los. P< =  0.01 = 1% los and ***,** and * indicate the estimated coefficient is significantly different from zero at the 1,5 and 10 percent levels, respectively.

However, female students feel less confident in their reconsideration of career commitment and occupational mobility, which implies that they feel more disappointed if their first choice of career does not work and feel less prepared for alternative career plans compared to men. The OECD (2015) states that girls tend to have more ambitious but unrealistic expectations about their career goals and may be more disappointed if they can’t achieve those goals.

The focus group shed light on these findings. With respect to the first finding, respondents emphasised the availability of more scholarships for female students in STEM, the fact that female students had to undergo more hardships to get into STEM as well as the fact that STEM disciplines were more focused as compared to non-STEM disciplines. The following quotations illustrate these:

I think this result might be attributed to the fact that there is a much higher percentage of scholarships for female students. (STEM 2, Male student)

Women are more confident in where they are going, especially in STEM as they know what they are getting into and what they want from the choice they have made. (Non-STEM 1, Female Student)

This has got to do with the fact that women have had to face problems before getting to university and they are more confident because of that. Their experiences in high school, experiences at home, with family and friends have made them stronger. (Non-STEM 2, Female Student)

This was consistent with the notion of Ertl et al. (2017) who suggest gender stereotypes negatively relate to STEM students’ self-concept and because women who study STEM subjects have likely already overcome barriers in school and family, they may be less affected by these influences or views after choosing a less-than-typical career path.

With respect to the second finding, the focus group participants presented mixed responses with some disagreeing with the quantitative findings (STEM 1 and 2) while others agreed (Non-STEM 1).

I don’t agree personally. If I don’t do engineering, I feel I am screwed – although I know other people think that guys can go into trades, become a handyman. So maybe that could be it. I could become an electrician or carpenter or a different tradesman – women do not go into those roles, so the perception of males would be that you can do that. But maybe women don’t think that way. There is a clear psychological difference between males and females. Males can just accept doing something first and then think about how to do it. Females think about it first and then do it. (STEM 1, Male student)

One of the factors would be theory and simulation are much more male dominated than industry and practical (the four streams in physics I mentioned earlier). People in theory and simulation can swap between the two – If I cannot do theory, I’ll do simulation. In areas of the discipline wherein there is more flexibility, there tends to be greater male dominance. (STEM 2, Male student)

How men think differently than women – maybe there is some psychology behind that. (Non-STEM 1, Female student)

Table 5 presents the gender mean score differences in the assessed constructs among non-STEM participants. Similar to STEM students, female students demonstrated higher confidence in career commitment and relevance of learning than their male counterparts, while male students reported higher scores for reconsideration of commitment and occupational mobility in non-STEM fields.

Table 5. Gendered mean differences in confidence among non-STEM students (n = 7353).

Constructs	Female (n = 4363) Mean (s.d.)	Male (n = 2990) Mean (s.d.)	Mean difference	p value
1. Career Identity and commitment	3.845 (.786)	3.784 (.800)	.061***	0.001
2. Reconsideration of career commitment	2.558 (1.023)	2.733 (.971)	-.174	0.000
3. Willingness and ability to learn	4.647 (.795)	4.606 (.780)	.040**	0.031
4. Relevance of learning	4.065 (.560)	4.024 (.573)	.040***	0.002
5. Career exploration & awareness	7.844 (1.447)	7.777 (1.424)	.066**	0.049
6. Occupational mobility	7.281 (1.658)	7.396 (1.552)	-.115	0.002

Notes: The mean difference is (female mean score – male mean score). +ve value means female score > male score which indicates, on average, females are more confident in that literacy skill. P< =  0.1 = 10% los, p< =  0.05 = 5% los. P< =  0.01 = 1% los and ***,** and * indicate the estimated coefficient is significantly different from zero at the 1,5 and 10 percent levels, respectively.

The findings from the focus group session largely supported the quantitative results. One student (Non-STEM 1, Female student) attributed the varied confidence levels around commitment and occupational mobility to the fact that many STEM programmes are accredited pathways.

STEM students have clearer pathways which are clearly laid out from the time they enter the course. With non-STEM, for example business, you are always evolving as there are several career opportunities with different job titles. An engineer will become an engineer. (Non-STEM 1, Female student)

The student also attributed this to men being more confident in their penchant for careful planning and resilience, although this is not supported in the literature.

It comes down to resilience. Some have plans and backup plans. One of my male friends – he has six backup plans. It was confronting to me as I normally have just one backup plan. STEM could have less restrictions as they have fewer options. I mean STEM choices are straightforward. You have few options … With non-STEM there are wide range of options – different job titles; … with non-STEM, you can change careers without having to get another degree. (Non-STEM 1)

Table 6 reports the mean score differences in confidence level of women in STEM versus non-STEM. The findings indicate that women in STEM reported higher confidence in all the indicators (with exception of career commitment 2 and occupational mobility) compared to women in non-STEM. This indicates women in STEM are more confident than women in other fields of study.

Table 6. Mean score differences in confidence levels of women in STEM vs. women in Non-STEM (n = 7157).

Constructs	Female (n = 2,794) Mean (s.d.)	Female (n = 4,363) Mean (s.d.)	Mean difference	p value
	STEM	Non-STEM
1. Career Identity and commitment	3.987 (.738)	3.845 (.786)	.141***	0.000
2. Reconsideration of career commitment	2.485 (1.061)	2.558 (.1.023)	-.073	0.003
3. Willingness and ability to learn	4.755 (.771)	4.647 (.795)	.108***	0.000
4. Relevance of learning	4.219 (.514)	4.065 (.560)	.153***	0.000
5. Career exploration & awareness	8.019 (1.372)	7.844 (1.447)	.174***	0.000
6. Occupational mobility	7.341 (1.628)	7.281 (1.658)	.059	0.134

Notes: The mean difference is (women in STEM – women in nonSTEM). +ve value means female score in STEM > female score in nonSTEM which indicates, on average, STEM females are more confident than nonSTEM females. p< =  0.1 = 10% los, p< = .05 = 5% los. p< = .01 = 1% los and ***, ** and * indicate the estimated coefficient is significantly different from zero at the 1,5 and 10 percent levels, respectively.

The focus group provided support for these findings and attributed the higher confidence levels in females being necessary due to the general difficulty for females to get into STEM given its relatively historic male dominance.

To be a woman in STEM, they have to be quite strong. There is a special something about them and they believe they are destined to do great things. There is a good mix of gifted girls in my engineering classes. (STEM 1, Male student)

I think that there is a distinct lack of diversity in STEM. Women in STEM have unique perspectives and they are able to contribute. (STEM 2, Male student)

Students know what they want in STEM and they need to fight hard to progress in the field. They already know what they are getting into beforehand. I know really smart women who never went into that area as they did not have the passion to go in and fight for it. Some women settle for their careers and go into non-STEM. (Non-STEM 1, Female student)

What drives those women towards STEM industries – they have passion for it, a motivation to go against the odds. With Non-STEM, they are just exploring, they don’t know if they love the subject yet when they enter university. (Non-STEM 2, Female student)

These views are supported by Ertl et al. (2017) who posit that we can assume that women who have chosen to be in fields with low women representation have already successfully overcome the barriers in school and family. At the same time, however, they claim that gender stereotypes are so strong and can still affect the women who have chosen these fields of studies. The current research extends Ertl et al. (2017) by considering not just academic self-concept but perceived employability.

Table 7 reports the gender wise mean score difference across the four STEM disciplines. Female students in the natural and physical sciences and IT were more confident in their willingness and ability to learn and relevance of learning (natural and physical sciences). Women in health were less confident in their reconsideration of career commitment and occupational mobility as compared to their male counterparts. These findings were similar to the findings discussed earlier, with the focus group confirming support for them.

Table 7. Area of study: Gender wise mean score differences in confidence levels (n = 5355).

Constructs	Gender Mean Difference – STEM
	Natural & Physical Sciences	Information Technology	Engineering & Related Technology	Health
1. Career Identity and commitment	.119 (p = 0.128)	.067 (p = 0.182)	.090 (p = 0.343)	.052 (p = 0.100)
2. Reconsideration of career commitment	-.126 (p = 0.198)	.126 (p = 0.060)	.166 (p = 0.215)	-.211*** (p = 0.000)
3. Willingness and ability to learn	.186** (p = 0.031)	.107** (p = 0.042)	.155 (p = 0.123)	.004 (p = 0.881)
4. Relevance of learning	.183*** (p = 0.001)	-.032 (p = 0.383)	.114 (p = 0.114)	.032 (p = 0.128)
5. Career exploration & awareness	.299 (p = 0.063)	.157 (p = 0.074)	.181 (p = 0.334)	-.056 (p = 0.335)
6. Occupational mobility	.110 (p = 0.544)	.112 (p = 0.282)	.045 (p = 0.825)	-.259*** (p = .000)

Notes: The mean difference is (female mean score – male mean score). +ve value means female score > male score which indicates, on average, females are more confident in that literacy skill. p< = .1 = 10% los, p< =  0.05 = 5% los. P< =  0.01 = 1% los and ***,** and * indicate the estimated coefficient is significantly different from zero at the 1,5 and 10 percent levels, respectively.

This study explored the efficacy beliefs of students in relation to perceived employability in STEM. Career and study efficacy emerge as important for two primary reasons. First, students with higher efficacy beliefs in relation to future employability are likely to demonstrate higher levels of self-determination, adopt a positive attitude to work, consider broader career choices, and pursue more opportunities (Berntson & Marklund, 2007; Kaleva et al., 2019). Second, assessing perceived employability affords students opportunities to compare their self-reports with those of peers and other stakeholders, potentially bringing into view areas of both agreement and disagreement (Donald et al., 2019). The gendered differences identified in this study emphasise the importance of perceived employability in both student decision making and student performance, with the development of efficacy beliefs crucial to the inner-value capitals (Baruch & Peiperl, 2000) on which study and career success depend, particularly for minority groups such as women in STEM.

Contributions

This paper makes several contributions to the literature which consequently have implications for practice. Firstly, the paper has demonstrated gender differences in perceived employability of STEM students during university studies. Because female students are less confident in reconsidering career commitment and occupational mobility than males, it is vital to create an intentional pedagogy of employability for STEM students where they can explore their identities and aspirations and develop personally and professionally while at university (Bennett et al., 2021a). Educators’ behaviour can support students’ interest in STEM fields (Ertl et al., 2017), by engaging in meaningful discussion about industry challenges and student concerns so as to help them navigate uncertain futures (Bennett et al., 2020b).

Secondly, to accelerate the STEM workforce in practice, stakeholders including government and industry need to work together to improve high quality education and training programmes in STEM (PCAST, 2020). Further, equal career opportunities must be created for men and women, which includes addressing socio-cultural norms in STEM through having gender inclusive policies and creation of a safe working environment (van Veelen et al., 2019).

Thirdly, all stakeholders, including the educators and higher education institutions, need to realise women’s potential in STEM fields of study. Low confidence among girls in their abilities to do mathematics can hold them back (OECD, 2015) and the learning environments and stereotypes, if not appropriate, can undermine their confidence further. Using a social cognitive framework, the use of self-assessment tools might help girls to question their low self-beliefs while encouraging educators to create opportunities for growth and reflection within supportive communities of peers and advocates.

Conclusion

STEM skills are in increasing demand and many organisations will become STEM, or STEM-aligned organisations (Australian Academy of Science, 2019). Thus, there will be continued calls to address gender equity gaps in STEM fields of study and workplaces. Findings from this study underpin the importance of understanding the impact and influence of gender on the perceived employability of STEM students. Neglecting these nuances risks increasing both the underrepresentation of women in STEM and the leaky pipeline. As women at all levels in STEM education and careers face barriers, providing girls and women an environment in which to progress is the shared responsibility of government, the education sector, and industry.

Acknowledgements

To be added post-review

Disclosure statement

No potential conflict of interest was reported by the author(s).

Appendix

Appendix 1: Validated instrument

Table

Appendix 2: Final factor structure

Table

[The authors will provide Appendix 1 as a URL, linking readers with the full instrument.]