Motivation plays a central role in students’ academic and career choices, and over the years, several prominent theories have emerged to explain how and why learners engage with particular fields (Gopalan et al., 2017). Expectancy-Value Theory (EVT), one of the most widely applied frameworks in STEM education, was originally developed by Atkinson and later expanded by Eccles and Wigfield, as noted by Dhanapala (2024). EVT posits that students’ motivation is shaped by their expectations for success and the subjective value they assign to a task, divided into intrinsic value, utility value, attainment value, and perceived cost (Eccles and Wigfield, 2020). This theory has been instrumental in explaining gender differences in academic and career choices by highlighting how cultural and social factors shape students’ expectations and values.

Another influential framework is Self-Determination Theory (SDT), developed by Deci and Ryan, which distinguishes between intrinsic and extrinsic motivation and emphasizes the fulfillment of three basic psychological needs; autonomy, competence, and relatedness, as shown in Howard et al. (2021). SDT has been widely used in educational research to explore how learning environments affect students’ engagement, persistence, and wellbeing (Abzhanova et al., 2024). It is particularly useful in understanding how autonomy-supportive environments can promote deeper motivation and persistence, especially in challenging fields such as engineering.

Social Cognitive Career Theory (SCCT), developed by Lent, Brown, and Hackett, builds directly on Bandura’s Social Cognitive Theory, which posits that learning and behavior are shaped by the dynamic interplay of personal factors, environmental influences, and behavior itself, a process Bandura termed reciprocal determinism (Amri et al., 2025). As noted by Grigg et al. (2018) the of SCCT is the concept of self-efficacy, or one’s belief in their ability to succeed in specific situations. SCCT applies this to academic and career contexts, suggesting that self-efficacy beliefs, together with outcome expectations and personal goals, strongly influence an individual’s career interests, choices, and persistence (Lent and Brown, 2019). Importantly, Li and Liu (2025) highlighted that SCCT also incorporates the role of contextual supports and barriers, such as access to role models, encouragement from teachers or family, and structural constraints, into its framework. In STEM education, and particularly in engineering, SCCT has been used to explain how confidence in one’s technical abilities, combined with expectations of positive career outcomes, shapes persistence (Costa et al., 2025). However, while SCCT acknowledges social and environmental influences, it still underrepresents deep cultural factors, gender norms, and the structural inequities that may shape the self-efficacy and goals of women in male-dominated fields like engineering (Li and Liu, 2025).

Beyond these three foundational models, other theories have contributed to understanding academic motivation. Achievement Goal Theory, as described in Williams (2024), differentiates between mastery goals, emphasizing learning and self-improvement, and performance goals, emphasizing surpassing others, both of which shape persistence and resilience in STEM. Attribution Theory explores how students interpret academic successes or failures, such as whether they attribute outcomes to ability, effort, or external factors (Hamm et al., 2017). This theory is relevant to understanding phenomena such as impostor syndrome, particularly among high-performing students who may undervalue their success. Interest Theory and Science Identity Theory emphasize the role of personal interest and identification with a scientific community in sustaining long-term engagement, particularly in fields where individuals may feel marginalized (Lockhart et al., 2022).

Additionally, Social Learning Theory highlights the importance of observing role models and peer influences, reinforcing the value of mentorship in shaping STEM motivation (Koutroubas and Galanakis, 2022). Expectancy Theory, often applied in organizational contexts, examines how individuals make choices based on expected outcomes and perceived effort-reward tradeoffs, which can relate to perceptions of job security and financial return in engineering (Fang, 2023). Finally, Self-Regulated Learning Theory integrates motivation with metacognitive skills including goal-setting, time management, and persistence, which are critical for success in rigorous academic environments such as engineering (Brenner, 2022).

While these theories provide strong foundations for understanding educational motivation, they also exhibit significant limitations, especially when applied to female students in engineering in non-Western settings (Narjes et al., 2025). Most dominant educational motivation theories were developed in Western, individualistic contexts, emphasizing personal agency, cognitive beliefs, and internal goals, while underrepresenting the influence of cultural values, family expectations, gender norms, and institutional prestige (Murayama and von Keyserlingk, 2025). They tend to treat motivation as an internal, rational process and often overlook the structural and sociocultural forces that shape educational decisions, particularly for women navigating gendered environments. As such, while these theories remain relevant, they are insufficient on their own to capture the complex and context-dependent motivations of female students in engineering education in the UAE. Therefore, this study will develop a Contextualized Female Engineering Motivation Model (CFEMM), extending foundational motivation theories through the integration of psychological, sociocultural, institutional, and national development factors to more accurately reflect the lived experiences of women in engineering in the UAE.

Building this model first requires situating the discussion within the broader global landscape of women’s participation in STEM, as understanding international patterns and challenges provides essential context for examining motivational dynamics.

The global underrepresentation of women in STEM persists despite decades of international policy efforts and institutional initiatives (Mattiello et al., 2025). According to the World Economic Forum’s Global Gender Gap Report 2024, women occupy only 28.2% of STEM roles worldwide, and their presence in senior leadership positions in these fields is considerably lower, reinforcing a long-standing “leaky pipeline” phenomenon (Gopalan et al., 2017). In particular, Dhanapala (2024) reported that women comprise approximately one-third of researchers in the physical sciences globally. Their projections suggest that, if current trends persist, full gender parity in these disciplines may not be reached until well into the next century, highlighting enduring structural barriers to equitable participation in scientific research.

Moreover, a bibliometric study conducted by Eccles and Wigfield (2020), which examined over 80 million scholarly publications spanning from 1975 to 2020, found that women represent less than 25% of top-ranked authorships across 19 academic disciplines. The study also revealed that in highly technical fields such as physics, geology, and mathematics, women’s representation can be as low as 14%. These findings underscore that gender disparities in STEM extend beyond enrollment and employment, affecting women’s visibility, influence, and opportunities for career advancement.

Much of this underrepresentation can be attributed to structural and cultural barriers. According to Howard et al. (2021), early gender socialization, along with culturally reinforced beliefs about ability and suitability, shapes students’ academic choices long before they reach university. According to Abzhanova et al. (2024), early gender socialization, together with culturally embedded perceptions of ability and appropriateness, significantly influences students’ academic choices before university. More recent work by Amri et al. (2025) provides evidence that implicit bias continues to affect hiring decisions, promotion rates, and leadership representation in engineering industries. This is supported by Grigg et al. (2018), where the authors highlight that these factors foster exclusionary academic environments in which female students, despite equal or superior performance, experience lower self-confidence, a diminished sense of belonging, and reduced academic satisfaction.

In the Middle East and North Africa (MENA) region, these global patterns are further shaped by strong cultural and familial expectations. For instance, Lent and Brown (2019) noted that traditional gender norms in the Gulf often cast engineering as incompatible with women’s roles in society, even as governments push for broader female participation in education. Although female enrollment in STEM programs across the MENA region has significantly increased, often exceeding participation rates in Western countries, the existing literature remains limited in scope. There is a critical need for scholarly inquiry that moves beyond enrollment statistics to investigate the underlying factors motivating women’s engagement in STEM and to critically examine their lived experiences within these fields (Li and Liu, 2025). As highlighted by Costa et al. (2025), the MENA region presents a “paradox of participation,” whereby formal inclusion through increased female STEM enrollment has not been matched by equitable representation in research, leadership, or professional advancement.

While challenges remain, the World Economic Forum (Li and Liu, 2025) highlights that female representation in Artificial Intelligence (AI) engineering roles has more than doubled since 2016. Despite persistent gender imbalances, this statistic is cited as evidence of incremental progress in gender equality within emerging technological fields. Moreover, Li and Liu (2025) also found through LinkedIn’s workforce data that women now hold over 42% of jobs globally, though only 31.7% of leadership positions. Furthermore, Williams (2024) highlighted that industry-specific reports, such as Tech Briefs, also show modest growth in women’s representation in engineering roles, especially in North America and Western Europe. Nonetheless, these gains are uneven across disciplines and regions and often fail to address the deeper psychosocial and institutional mechanisms that deter women from persisting and advancing in technical fields.

The UAE offers a particularly compelling context within the MENA region, as it actively challenges traditional gender norms through national initiatives aimed at advancing women in STEM fields (Hamm et al., 2017). However, Lockhart et al. (2022) caution that this progress may be superficial in some areas. Their cross-sectional study of 165 female STEM students and faculty found that 75% of respondents perceived ongoing gender inequality in labs and classrooms, often citing a disconnect between policy rhetoric and day-to-day institutional practices. Given these persistent challenges, there is a critical need to examine how gendered academic environments in the UAE affect female students’ confidence, sense of belonging, and academic satisfaction, factors often overlooked despite increasing female participation in STEM. Addressing these challenges begins with understanding the drivers of women’s engagement with STEM, insights that form the basis for the following section on motivational factors.

Understanding the underpinning factors that motivate women to pursue STEM disciplines is critical to addressing gender disparities in these fields. A growing body of research suggests that women’s decisions to enter STEM are influenced by a complex interplay of personal interests, social expectations, cultural norms, and perceived career opportunities (Mansour, 2025). Historically, as illustrated by Eccles and Wigfield (2024), Expectancy-Value Theory has provided a foundational framework for examining these decisions, positing that academic and career choices are shaped by individuals’ expectations of success and the subjective value they assign to particular fields. For women in STEM, this involves weighing personal interest against societal expectations and perceived barriers to entry and success. In particular, Phillips (2024) discovered that internal drivers, such as intellectual curiosity, personal interest in problem-solving, and a desire to make societal contributions, often lead women toward STEM fields. Additionally, Felgueira et al. (2024) recognized that career aspirations linked to innovation, technology, and leadership further reinforce these choices. However, Siwale and Mwalemba (2023) discovered that external factors, including job security, financial independence, family expectations, and social perceptions of prestige, are frequently cited as additional influences on women’s decisions to enter technical disciplines.

Despite these motivations, many female engineering students experience the impostor phenomenon, a persistent disconnect between their actual academic achievements and their self-perceived competence (Vicent, 2025). Studies, including (Whitcomb et al., 2020; Liberatore and Wagner, 2022; Lee et al., 2024), consistently show that women in engineering perform as well as or better than their male peers, yet often report lower levels of self-efficacy. This confidence gap can lead to dropouts, unless mitigated by supportive learning environments, mentoring relationships, and recognition from faculty and peers (Lübeck et al., 2025). Additionally, Beroíza-Valenzuela and Salas-Guzmán (2025) acknowledged that environments that affirm women’s abilities play a crucial role in reinforcing persistence and fostering positive academic identity.

Belonging is another critical component of women’s motivation in STEM (Preuß et al., 2025). It was highlighted by González-Pérez et al. (2022) that when women perceive engineering as culturally framed as a male domain, they are more likely to experience feelings of isolation and self-doubt, which can undermine their sense of fit and reduce their willingness to continue in the field. A strong sense of belonging is directly linked to retention and long-term engagement in STEM programs.

Mentorship and the presence of role models also play a central role in sustaining women’s motivation and persistence, as noted by García-Silva et al. (2025). In particular, Wu et al. (2022) found that female-to-female mentoring significantly reduces dropout rates in engineering, even when such interventions do not directly impact grades. Mentors help reinforce women’s STEM identities, provide social and academic support, and create pathways to leadership opportunities.

Social and cultural influences further shape women’s STEM experiences (Cernadas et al., 2025). As emphasized by Ferreira et al. (2025), family expectations, cultural norms, and societal narratives about gender-appropriate careers often steer women’s educational decisions. Furthermore, Vossoughi et al. (2025) found that experiences with stereotypes, unconscious bias, and subtle negative interactions may affect motivation, particularly for women from underrepresented backgrounds. As such, these situations and experiences may create additional psychological stress, particularly when women must navigate environments that conflict with their personal or cultural identities.

On the other hand, Beroíza-Valenzuela (2025) recognized that resilience and coping strategies are crucial for women who persist in engineering despite these challenges. According to recent studies, such as Ojong and Kareem (2025), Lekhak (2023), and Pineda et al. (2025), many women rely on passion for the field, self-efficacy, and adaptive strategies such as peer support networks, academic resource utilization, and seeking mentorship to overcome situational barriers and hidden curriculum pressures. Such resilience enables women to manage setbacks and maintain motivation throughout their academic and professional journeys (Bustamante-Mora et al., 2025).

Institutional reputation has also emerged as a significant factor influencing women’s decisions to pursue engineering education (Torres et al., 2025). It was suggested by Moridnejad et al. (2022) that students are more likely to enroll in programs perceived as prestigious, well-resourced, or highly ranked, believing these attributes correlate with better career outcomes and professional credibility. For women entering male-dominated disciplines such as engineering, Gille et al. (2022) found that the reputation of a university or specific program can offer a sense of validation and legitimacy, potentially offsetting concerns about gender bias or future workplace discrimination. A well-regarded institution may also signal a stronger support infrastructure, further reinforcing women’s confidence in their academic and professional trajectory.

Another motivational factor increasingly recognized in recent literature is the desire to contribute to national development and societal progress through engineering (Ramos-Gavilán et al., 2024). On the other hand, Vaez Ghaemi et al. (2024) emphasizes that female students view engineering as a way to contribute to solving practical problems in fields like infrastructure, sustainability, healthcare, and technology. As Barsoum (2021) demonstrates, the drive to serve the public good and uphold civic values plays a significant role in motivating many women in the field. Engineering is seen not only as a technical pursuit but as a means of shaping societal advancement, aligning personal goals with broader national or global agendas such as the Sustainable Development Goals (SDGs) (Ramirez-Mendoza et al., 2020).

Financial support mechanisms, such as scholarships, grants, and merit-based funding, play a crucial role in shaping women’s motivation to enter and persist in STEM (Petean and Rincon, 2024). As noted by Lasekan et al. (2024), competitive scholarships help reduce financial obstacles while also boosting students’ self-esteem and academic involvement. As demonstrated by Helman et al. (2020), transparent and inclusive funding structures not only provide institutional recognition and validation but also signal potential and belonging, factors that reinforce many women’s commitment to succeed in rigorous STEM programs.

Finally, structured interventions that promote leadership identity, community-building, and professional recognition have been shown to enhance women’s long-term aspirations in STEM (Rondon Pereira et al., 2025). According to Melendez-Anzures et al. (2025), programs that actively support women as future leaders in STEM fields help them succeed academically and motivate them to stay in the field after graduation. These findings underscore the importance of creating inclusive academic environments that nurture both competence and belonging, while addressing systemic barriers that hinder women’s full participation in STEM.

However, despite the breadth of existing research, these motivational factors remain insufficiently understood in diverse educational contexts (Masjutina and Stearns, 2025). As identified by Alzaabi et al. (2021), much of the existing literature focuses on Western contexts, creating a gap in understanding how these factors operate across diverse cultural, institutional, and social settings. Further research is needed to examine how women’s motivations, self-efficacy, belonging, mentorship experiences, and resilience interact throughout their academic journeys in STEM.

In response to this gap, the present study examines how motivational indicators influence women’s participation and persistence in STEM within contexts of educational expansion and shifting societal expectations. To provide a clear conceptual foundation for this investigation, the principal motivational factors identified in the existing literature are synthesized and presented in Table 1. These indicators are conceptually grouped into intrinsic and extrinsic motivational categories, reflecting internal drivers and external influences commonly discussed in STEM motivation research. This classification directly informs the methodological approach of the study by guiding the development of the qualitative interview protocol and the subsequent survey instrument, while also positioning the proposed CFEMM model within the broader literature on women’s motivation in STEM.

While gender disparities in STEM have been widely studied, there is limited research that captures how these disparities manifest within specific institutional and cultural settings, especially in countries experiencing rapid modernization and educational expansion, such as the UAE (Dickson and Al Harthi, 2023).

In particular, AUS presents a distinctive and compelling case for such exploration (Kjerfve, 2017). Founded in 1997, AUS is a leading higher education institution in the Gulf region, known for its rigorous American-style liberal arts enriched curriculum, strong emphasis on engineering and architecture, and commitment to diversity and academic excellence (American University of Sharjah, n.d.a). It is accredited both in the UAE and by the U.S.-based Commission on Higher Education of the Middle States Association, further enhancing its global academic credibility (Commission for Academic Accreditation (CAA), 2021). According to its official statistics, the American University of Sharjah (AUS) hosts students from over 70 nationalities, fostering a multicultural and co-educational environment that supports cross-cultural dialogue and diversity in learning (American University of Sharjah, n.d.a). This creates a valuable context to examine how women navigate STEM education in internationally oriented classrooms while still embedded within the cultural norms and expectations of the broader UAE society (Dickson et al., 2023).

AUS is recognized for its excellence in engineering education through its College of Engineering (CEN), the largest college at the university, with over 2,500 undergraduate and 350 graduate students representing more than 70 nationalities. Accredited by the UAE Ministry of Higher Education & Scientific Research (MOHESR) and USA-based Accreditation Board for Engineering and Technology (ABET), the college offers 10 undergraduate degree programs, 10 master’s programs, and 4 PhD programs. Female participation is strong, particularly in STEM disciplines traditionally dominated by men, with women comprising 36% of undergraduates and 46% of graduate students. CEN is staffed by 104 faculty members, 26 lab instructors, and 21 administrative staff. Moreover, it supports hands-on learning through 76 advanced technical labs housed in a state-of-the-art 38,000 m² facility. Ranked as number 345 globally for Engineering and Technology in the QS World University Rankings by Subject 2025, AUS emphasizes diversity, innovation, and mentorship, having awarded over 9,000 engineering degrees to date (American University of Sharjah, n.d.b). Through active student organizations, mentoring, and inclusive policies, the college fosters an environment where women are encouraged and supported to thrive in engineering fields.

Given these qualities, AUS, and CEN in particular, offers a strong case for examining the factors shaping women’s involvement, success, and persistence in engineering. Its blend of global standards, cultural diversity, and commitment to inclusion provides an ideal setting to study how local and international influences intersect. This focus also informs the development of the CFEMM, a framework that combines established motivation theories with context-specific cultural, institutional, and psychological factors. In this study, CFEMM is applied to analyze how intrinsic and extrinsic motivation, sense of belonging, self-efficacy, and the surrounding sociocultural and institutional environment collectively influence female engineering students at AUS, while also serving as a model for broader application in other educational contexts.

The study population consisted of undergraduate female engineering students enrolled at the American University of Sharjah (AUS). Both phases of the study employed criterion-based purposive sampling, a non-probability method in which participants are selected according to predefined inclusion criteria relevant to the research objectives (Campbell et al., 2020). In the qualitative phase, participants were chosen to ensure variation across engineering majors, including civil, mechanical, electrical, chemical, and computer engineering, and academic years (ranging from freshman to senior level).

In this phase, a subset of high-achieving female engineering students was intentionally included to support in-depth exploration of motivational factors within the institutional context under study. As the aim of this phase was to identify information-rich insights related to persistence, self-efficacy, and engagement in engineering, high-achieving students were well-positioned to reflect on how contextual supports and institutional structures influenced their motivation. This selective sampling approach aligns with the study’s exploratory and model-development focus. Nevertheless, this choice may foreground resilience-oriented motivational themes; accordingly, qualitative findings are interpreted as illustrative rather than representative.

In the quantitative phase, the same criterion-based purposive approach was used, whereby only female students enrolled in AUS engineering programs were invited to participate through classroom announcements and university email. This approach ensured that all participants were directly relevant to the study’s focus while allowing for broader participation across engineering programs and academic levels. The quantitative survey yielded 180 valid responses from undergraduate female engineering students. Based on institutional enrollment data indicating that women comprise approximately 36% of the 2,568 undergraduate engineering students at AUS (≈924 students), this corresponds to an estimated response rate of about 20%.

The semi-structured interview protocol was developed based on the motivational indicators synthesized from the literature and presented in Table 1. To translate this broad set of indicators into an interview framework, the indicators were first conceptually grouped into intrinsic and extrinsic motivational dimensions. Intrinsic dimensions captured internal drivers such as intellectual curiosity, enjoyment of problem-solving, self-efficacy, sense of belonging, and the desire to contribute to society. Extrinsic dimensions reflected external influences, including job security, financial independence, institutional reputation, scholarships, and social or family encouragement. These dimensions were selected because they were consistently used across prior studies, aligned with established motivation theories, and suitable for in-depth qualitative exploration within a semi-structured interview format.

The combination of qualitative interviews and quantitative surveys ensured both depth and breadth in understanding the motivations of female students for pursuing engineering. Prior to data collection, the interview guide was pilot-tested with two female engineering students who met the inclusion criteria but were not part of the final sample, leading to minor refinements in question wording and sequencing. Semi-structured interviews were then conducted online with 10 high-achieving undergraduate female engineering students at CEN, AUS, during the summer period (July to August). Each interview lasted approximately 40–60 min. Guided by the intrinsic and extrinsic motivational dimensions outlined above, the interviews explored students’ personal experiences and perceptions related to their decision to pursue engineering. The open-ended format allowed participants to elaborate freely on factors they considered most influential, with the aim of identifying salient motivational indicators for subsequent quantitative assessment.

Building on the qualitative findings, a structured survey was developed to operationalize the motivational indicators identified through the interviews. Survey items were newly developed for this study by synthesizing motivational indicators identified in the literature with emergent qualitative themes specific to the institutional context. To establish content validity, the draft survey was reviewed by faculty members with expertise in engineering education and STEM motivation, leading to minor refinements in wording and clarity. The instrument was pilot-tested with a small group of female engineering students who met the inclusion criteria but were not part of the final sample. All survey items were measured using a 5-point Likert scale.

Qualitative data were analyzed using thematic analysis. Interview transcripts were reviewed multiple times to achieve familiarization, followed by open coding to identify recurring ideas related to motivational drivers. Codes were then grouped into broader themes corresponding to intrinsic, extrinsic, and social influences. The most salient themes informed the refinement of motivational indicators and the final survey instrument used in the quantitative phase.

The quantitative data collected from the survey are analyzed using a three-step statistical approach: Exploratory Factor Analysis (EFA), Confirmatory Factor Analysis (CFA), and Structural Equation Modeling (SEM), which will inform the development of the CFEMM. The overview of the analytical approach is shown in Figure 3. All of the data analysis techniques will be conducted using Statistical Analysis System (SAS).

As the qualitative section collected the viewpoints of female engineering students at AUS, EFA is essential to understand the correlation between the indicators and to confirm the grouping identified through the literature review. EFA is a multivariate method used to identify and group key variables into meaningful factors that explain underlying structures across various fields, including social, health, and economic sciences (Sürücü et al., 2024). This allows the development of a model that is relevant to the context of the case study.

Prior to conducting factor and structural analyses, preliminary diagnostics were performed to assess key statistical assumptions, including item distributions, linearity, and multicollinearity. While no severe violations were observed, minor deviations from multivariate normality were identified, which is common in survey-based motivation research. Accordingly, Generalized Least Squares (GLS) estimation was employed, as it is less sensitive to normality assumptions than Maximum Likelihood estimation and is appropriate for exploratory and theory-building applications with moderate sample sizes (Du and Bentler, 2022).

Sequentially, CFA was conducted to validate the underlying structure of the motivational constructs identified and validated through the EFA. This multivariate statistical technique is used to test whether the observed survey items (indicators) accurately and consistently measure the latent variables (constructs) they are intended to represent (Khojasteh, 2022). Several model fit indices were calculated to evaluate the adequacy of the measurement model. These included the Goodness-of-Fit Index (GFI) and Comparative Fit Index (CFI), both of which compare the specified model to a baseline model, with values above 0.90 indicating an acceptable fit and values above 0.95 indicating an excellent fit (Goretzko et al., 2024). The Root Mean Square Error of Approximation (RMSEA) was also calculated, with values below 0.08 indicating a reasonable fit (Pavlov et al., 2021).

In this study, the CFA not only served to confirm the validity and reliability of the motivational constructs but also provided the empirical basis for structuring the CFEMM. By ensuring that each latent construct was measured accurately, the CFA results informed the SEM phase, where the relationships among these constructs were tested.

Furthermore, SEM is used to test and quantify the relationships between latent motivational constructs and their observed indicators. SEM combines features of factor analysis and multiple regression, allowing for the simultaneous estimation of measurement models and structural models (Hair, 2021). In this research, the SEM analysis will be grounded in the validated factor structure derived from the CFA, ensuring that reliable indicators represent each construct.

The structural model will be specified based on theoretical relationships identified in the first two stages of the analysis. This will include hypothesized causal paths and potential mediating effects. Model fit will be assessed using the same multiple indices used in CFA to ensure model fit and strength. The final SEM output will provide standardized path coefficients (β) indicating the magnitude and direction of each relationship. This comprehensive approach enables a nuanced understanding of how the motivational indicators interact to influence the academic and career choices of female engineering students. The final output of the SEM paths will inform the CFEMM.

The study adhered to the ethical standards outlined by the American University of Sharjah’s Institutional Review Board (IRB). The research protocol (including participant recruitment, consent procedures, and data handling) was reviewed and approved by the IRB under Protocol No. 25–105. Participants were fully informed about the purpose, scope, and procedures of the study prior to participation, and written informed consent was obtained. Anonymity and confidentiality were maintained by removing personally identifying information from all datasets and storing data securely in password-protected files accessible only to the research team. All data were used solely for academic purposes in accordance with the approved protocol.

Table 2 presents the demographic profile of the participants who took part in the semi-structured interviews. The sample includes 10 undergraduate female engineering students from AUS, representing a range of majors and academic levels. This diversity ensured that perspectives were drawn from different areas of specialization and stages of study, allowing for a more comprehensive understanding of students’ motivations and experiences in engineering.

The quantitative phase included 180 female engineering students from the College of Engineering (CEN) at the American University of Sharjah (AUS). Table 3 summarizes their demographic characteristics, including year of study, engineering major, and nationality. The distribution reflects broad representation across academic levels and disciplines, ensuring that the survey findings capture the diverse experiences of female engineering students within the population.

As shown in Table 3, the distribution across academic years is fairly balanced, with the largest share in the first year. The majors cover a broad range of engineering disciplines. Civil Engineering has relatively few respondents, as expected, given its smallest departmental enrollment. The “Other” category includes students from Intelligent Systems and Mechatronics Engineering, Computer Science, and Digital Construction and Management, which reflects the growing presence of emerging and interdisciplinary programs. The nationality percentages also indicate a diverse cohort, with Emirati, Jordanian, and Egyptian students making up the largest groups. Overall, the table provides a clear profile of the sample and shows meaningful variation in academic stage, field of study, and national background.

The semi-structured interviews with female engineering students at AUS revealed a rich mix of intrinsic, extrinsic, and cultural motivators shaping their academic journeys. A reflexive thematic analysis approach was employed, involving iterative open coding followed by the organization of codes into higher-order themes. Themes were refined through constant comparison across transcripts.

Before conducting the exploratory factor analysis, descriptive statistics were examined to provide context for the survey responses. Table 5 reports the item-level means and standard deviations for the 10 motivational indicators (ID1-ID10). Overall, the items show adequate variability and no evidence of extreme floor or ceiling effects, supporting their suitability for factor analysis.

The EFA examined the 10 indicators (ID1 to ID10) to identify their underlying latent structure using principal-axis factoring with Promax rotation. Two factors were retained based on the screen test and eigenvalue distribution, supporting a low-dimensional and interpretable solution. The rotated pattern matrix showed that ID2, ID4, ID5, ID6, ID7, ID8, and ID10 loaded primarily on Factor 1, while ID1, ID3, and ID9 loaded more strongly on Factor 2. Several indicators exhibited cross-loadings, reflecting conceptual overlap between intrinsic, social, and institutional influences. Given the exploratory nature of the study and the theory-informed development of CFEMM, items were retained and assigned to the factor on which they showed the highest absolute loading, guided by conceptual interpretability rather than strict loading thresholds. Overall, the two-factor model accounted for most of the shared variance and revealed a clear conceptual distinction between the two latent dimensions. The standardized regression coefficients of the indicators with respect to the factors are presented in Table 6.

Items loading on Factor 1 reflect a mix of personal aspiration and supportive social influences. These include family influence on the decision (ID2), belief that engineering offers promising career and financial prospects (ID4), the intention to challenge gender stereotypes (ID5), the influence of AUS’s engineering reputation (ID6), inspiration from female role models (ID7), a desire to contribute to national development (ID8), and alignment with long-term personal or professional goals (ID10). Together, these indicators suggest an Intrinsic and Empowerment Motivation factor reinforced by close social and institutional contexts.

On the other hand, Factor 2 groups items tied to individual interests and situational opportunities, including enjoyment of problem-solving and logical thinking (ID1), encouragement from teachers or mentors during high school (ID3), and attraction to scholarship or funding opportunities (ID9).

In summary, the two-factor solution distinguishes between Intrinsic and Empowerment Motivation (Factor 1) and External and Contextual Influence (Factor 2), reflecting both personal aspirations and the social dynamics underlying students’ decisions to pursue engineering.

Data validation was conducted prior to CFA to assess the reliability and validity of the collected indicators. The first step was to evaluate whether the sample size was sufficient. The Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy was used, with a threshold of 0.50. The data yielded a KMO value of 0.71, indicating a good level of sampling adequacy. The reliability of the data was assessed using Cronbach’s α and Composite Reliability (CR). The model validity was examined via CFA using GLS estimation. The results are summarized in Table 7.

The internal consistency of each dimension was assessed using Cronbach’s α and CR. The IEM construct recorded a Cronbach’s α of 0.534 and a CR of 0.703, indicating modest but acceptable reliability for exploratory purposes. In contrast, the ECI construct showed a Cronbach’s α of 0.322 and a CR of 0.341, reflecting low internal consistency. This low α value likely reflects the limited number of indicators and the conceptual breadth of the ECI construct. Given the exploratory and context-sensitive nature of the study, ECI is treated as an exploratory construct rather than a fully established scale, and reliability estimates are interpreted cautiously and primarily as indicative rather than confirmatory. Despite the modest reliability coefficients, all individual indicators within their respective constructs demonstrated positive associations, supporting their inclusion in the subsequent CFA.

Convergent and discriminant validity metrics, such as Average Variance Extracted (AVE), were not emphasized due to the exploratory nature of the measurement model and the context-driven development of the indicators. Instead, construct evaluation focused on factor structure, theoretical coherence, and overall model fit, which are appropriate for early-stage model development.

CFA was employed to validate the measurement structure of the motivation constructs identified in the literature and refined through qualitative analysis. The CFA evaluated whether the observed survey items load reliably on their intended latent variables. Based on the EFA results, items were reassigned to the factor on which they showed the strongest loading and recoded accordingly. The resulting indicator for factor grouping and recoding is reported in Table 8.

The CFA results indicate an acceptable model fit, as summarized in Table 9. The chi-square test was significant, χ²(32) = 82.27, p < 0.001, which is typical with samples of this size; therefore, greater emphasis is placed on descriptive fit indices. The Goodness of Fit Index (GFI = 0.91) and Adjusted GFI (AGFI = 0.84) meet recommended thresholds, indicating that the model reproduces the observed covariances adequately. The RMSEA of 0.09 reflects an acceptable level of approximation error. Overall, these indices support the adequacy of the proposed two-factor measurement model.

As shown in Table 10, the CFA standardized absolute path estimates show that all indicator loadings were statistically significant (p < 0.05) and within acceptable ranges. This confirms that the observed items appropriately represent their respective latent constructs. For the IEM factor, standardized estimates ranged from 0.37 (IEM4) to 0.76 (IEM7), indicating that all seven indicators make meaningful contributions to the construct, with IEM7 being the strongest indicator. For the ECI factor, standardized estimates ranged from 0.26 (ECI3) to 0.58 (ECI1), reflecting moderate but significant relationships between the indicators and their latent variable. These results demonstrate that both factors exhibit satisfactory convergent validity, with each indicator showing a statistically significant and theoretically consistent relationship to its underlying construct.

To investigate the impact of the latent motivational dimensions on the motivation of women in engineering (MWiE), SEM was constructed. In this model, IEM and ECI function as endogenous latent factors that predict the exogenous observed variable, MWiE. The following hypotheses were examined:

The SEM was fitted using PROC CALIS in SAS. The model fit indices are summarized in Table 11. The results show that the model achieves a GFI of 0.883 and an AGFI of 0.812, both exceeding the commonly accepted threshold of 0.80, indicating a good absolute fit. The RMSEA of 0.1 suggests a borderline but interpretable level of approximation error for an exploratory structural model.

Table 12 presents the SEM path estimates for the relationships among the two latent motivational constructs (IEM and ECI) and the outcome variable MWiE, which represents the overall motivation of women in engineering. The results show that ECI has a significant positive effect on MWiE (Estimate = 0.88394, p = 0.0071). This indicates that women whose motivations are shaped by external influences, such as family encouragement, scholarships, or institutional reputation, tend to report higher overall motivation toward studying engineering.

In contrast, IEM has a significant negative effect on MWiE (Estimate = −0.47942, p < 0.0001). This suggests that when women’s motivations are more internally driven, centered on personal goals, empowerment, or intrinsic interest, their overall MWiE score tends to be lower. This negative association does not indicate reduced interest in engineering. Rather, it suggests that the overall motivation measure (MWiE) is more closely aligned with externally reinforced and institutionally visible forms of motivation than with empowerment-oriented internal drivers within this specific institutional context.

Most measurement paths were significant at p < 0.05, with strong contributions from IEM4, IEM6, and IEM7 for the IEM factor, and from ECI1 and ECI2 for the ECI factor. A few indicators (IEM1, IEM5, and ECI3) were nonsignificant, consistent with their relatively low R-square values. Residual variances for the endogenous latent constructs (IEM and ECI) were also significant, reflecting unexplained variance that is typical in motivational models.

Overall, the SEM results identify that both hypotheses are significant. The findings show that women’s motivation toward engineering influences the two latent dimensions in distinct ways. Higher overall motivation is associated with a greater reliance on external and contextual influences, as well as lower levels of intrinsic and empowerment-based motivations. The structural paths are strong and significant, offering meaningful insight into how different motivational forces operate within the context of women in engineering.

For the measurement portion of the model, most IEM indicators were significant at p < 0.05, with the strongest standardized contributions coming from IEM4, IEM6, and IEM7. Two indicators, IEM1 and IEM5, were nonsignificant, consistent with their low R-square values. For the ECI construct, ECI1 and ECI2 were significant, whereas ECI3 was not, indicating variation in the strength with which each item reflects the underlying contextual influence dimension. These estimates align with the standardized variance results presented in the output. The final output of the statistical models is shown in Figure 4.