Technology spreads rapidly in the 21st century. It is possible to evaluate the 21st century as the age of knowledge, information and space (UN, 2021; UNESCO, 2023). There are new developments every day in all fields of science (Shelley and Kiray, 2018). Technological advancements contribute to the economic development of countries (OECD, 2024). Various sectors and working areas have emerged thanks to economic and technological progress (Zohuri, 2020). The 21st-century education model focuses on cultivating productive, inquisitive, critical thinking, and problem solving students. As technology advances in all societies, students’ fundamental life skills become compatible with the age of space and information (Alpaydın and Demirli, 2022; Paechter et al., 2020).

STEM (science, technology, engineering, and mathematics) provides opportunities for students to address problems from an interdisciplinary perspective (Freeman et al., 2019; Kennedy and Odell, 2024). It is aimed to provide knowledge and skills with a holistic educational approach covering the entire education process from preschool to higher education (Titrek et al., 2023; Yan et al., 2024). Different disciplines are brought together with STEM education. It becomes possible to achieve an inclusive education that supports quality learning, the application of existing knowledge to daily life, the development of life skills and high-level critical thinking. The STEM approach is important in terms of enabling the transformation of theoretical knowledge in the fields of science, technology, engineering and mathematics into practice and products (Kennedy and Odell, 2024; OECD, 2024; Sheth, 2023).

STEM education aims at economic progress (Hu and Guo, 2021; Yan et al., 2024). Countries are adopting various approaches to address gender divides. However, while 68% of countries globally have policies to support STEM education, only half of these policies specifically support girls and women. The United States has set a goal to increase the number of STEM bachelor’s degrees by one million over a 10-year period (Young et al., 2018). In Kenya, the Ministry of Education has organized STEM boot camps in schools, aiming to encourage and empower girls to pursue STEM studies and careers. In Namibia, it has mandated Goal 4 of the National Science and Technology Innovation Policy. It has improved gender inequality in STEM and the participation of women in science education. It has established programs that support the participation of girls as leaders and decision-makers in science careers. Zambia’s Technical Education, Vocational and Entrepreneurship Training Authority has established a platform that offers free digital skill courses targeted, among others, at women. In Bangladesh, the Eighth Five Year Plan 2020–25 has a dedicated section on STEM education for girls. During its G20 presidency, India launched the Tech Equity platform to empower women with digital literacy skills (UNESCO, 2024). Türkiye is one of the countries with the lowest rate of STEM graduates.

STEM fields are frequently promoted due to the high labour-market demand for graduates (Bybee, 2010). However, these fields also show gender disparities, with women under-represented in many STEM disciplines due to societal norms and educational biases. Progress in encouraging more women to pursue STEM-related fields has been slow. Across OECD countries, the proportion of female new entrants choosing to study STEM fields increased by less than 1 percentage point between 2015 and 2022 (OECD, 2024). It can be said that there is a strong positive correlation between individuals’ inclination towards STEM and their motivation (Master et al., 2017).

Clustering analysis was conducted for each subscale to classify attitude levels of female health sciences faculty students, considered disadvantaged in terms of STEM, towards scientific research. The results are presented in Table 4.

Table 4 indicates that the results of the clustering analysis for all scale scores are statistically significant (p < 0.05), indicating that the classification is interpretable. According to the clustering analysis results for the subscale scores of reluctance to help researchers, 42.77% of the participants (n = 68) have a low level of reluctance to help researchers, 53.46% (n = 85) have a moderate level, and 3.77% (n = 6) have a high level. According to the clustering analysis results for the subscale scores of negative attitudes towards research, 50.31% of the participants (n = 80) have a low level of negative attitude towards research, 47.80% (n = 76) have a moderate level, and 1.89% (n = 3) have a high level. Regarding the positive attitude towards research subscale scores, the clustering analysis results indicate that 8.81% of the participants (n = 14) have a low level of positive attitude towards research, 67.30% (n = 107) have a moderate level, and 23.90% (n = 38) have a high level. Regarding the positive attitude towards researchers subscale scores, the clustering analysis results indicate that 3.14% of the participants (n = 5) have a low level of positive attitude towards researchers, 27.67% (n = 44) have a moderate level, and 69.18% (n = 110) have a high level. In general, the majority of female students enrolled in the health sciences faculty have low to moderate levels of reluctance to help researchers and negative attitudes towards research, while having moderate levels of positive attitudes towards research and high levels of positive attitudes towards researchers.

Clustering analysis was conducted for each subscale to determine intelligence levels in the multiple intelligence subscales of female health sciences faculty students, considered disadvantaged. The results are presented in Table 5.

Table 5 indicates that the results of the clustering analysis for all scale scores are statistically significant (p < 0.05), indicating that the classification is interpretable. According to the clustering analysis results for the verbal–linguistic intelligence subscale scores, 3.77% of the participants (n = 6) have low verbal–linguistic intelligence level, 74.21% (n = 118) have a moderate level, and 22.01% (n = 35) have a high level. Regarding the logical-mathematical intelligence subscale scores, the clustering analysis results indicate that 8.81% of the participants (n = 14) have low logical-mathematical intelligence level, 73.58% (n = 117) have a moderate level, and 17.61% (n = 28) have a high level. The clustering analysis of the musical intelligence subscale scores reveal that 16.98% of the participants (n = 27) have low musical intelligence level, 67.30% (n = 107) have a moderate level, and 15.72% (n = 25) have a high level. Regarding the bodily-kinesthetic intelligence subscale scores, the clustering analysis indicates that 5.66% of the participants (n = 9) have low bodily-kinesthetic intelligence level, 75.47% (n = 120) have a moderate level, and 18.87% (n = 30) have a high level. For the spatial intelligence subscale scores, the clustering analysis results show that 17.61% of the participants (n = 28) have low spatial intelligence level, 71.07% (n = 113) have a moderate level, and 11.32% (n = 18) have a high level. Regarding the interpersonal intelligence subscale scores, the clustering analysis reveals that 3.77% of the participants (n = 6) have low interpersonal intelligence level, 42.77% (n = 68) have a moderate level, and 53.46% (n = 85) have a high level. The clustering analysis of intrapersonal intelligence subscale scores indicates that 5.03% of the participants (n = 8) have low intrapersonal intelligence level, 49.69% (n = 79) have a moderate level, and 45.28% (n = 72) have a high level. For the naturalist intelligence subscale scores, the clustering analysis results show that 6.29% of the participants (n = 10) have low naturalist intelligence level, 74.21% (n = 118) have a moderate level, and 19.50% (n = 31) have a high level. Overall, the majority of female students have moderate levels of verbal–linguistic, logical-mathematical, musical, bodily-kinesthetic, spatial, and naturalist intelligence, while their interpersonal and intrapersonal intelligence levels are moderate to high.

To determine the STEM value-expectancy assessment levels of female students, considered disadvantaged groups in terms of STEM, classification was performed through clustering analysis for scale scores. The results are presented in Table 6.

Table 6 indicates that the results of the clustering analysis for scale scores are significant (p < 0.05), indicating that the classification is interpretable. According to the clustering analysis results for STEM Value-Expectancy Assessment Scale scores, 6.92% of the participants (n = 11) have a low level of STEM value-expectancy assessment, 81.13% (n = 129) have a moderate level, and 11.95% (n = 19) have a high level.

To examine the correlation between scores of the Scale of Attitude Towards Scientific Research, the Multiple Intelligence Scale, and the STEM Value-Expectancy Assessment Scale, the Pearson correlation analysis was conducted. The results are presented in Table 7.

Correlation coefficients <0.20 are considered very weak, 0.20–0.39 is considered weak, 0.40–0.59 is considered moderate, 0.60–0.79 is considered strong and > 0.80 is considered very strong (Papageorgiou, 2022).

The study findings first examined whether female leadership candidates’ positive attitudes toward science positively affected their STEM motivation. It is then determined whether the multiple intelligence areas of female undergraduate students positively affect their STEM motivation. Table 7 shows that there is no significant correlation between the subscales of unwillingness to help researchers, negative attitudes towards research and logical-mathematical intelligence, musical intelligence, bodily-kinesthetic intelligence and spatial intelligence (p > 0.05). However, there are significant negative medium-level correlations between the subscales of unwillingness to help researchers and interpersonal intelligence, interpersonal intelligence (r = −0.363; r = −0.383) and a significant negative low-level correlation with naturalistic intelligence (r = −0.194) (p < 0.05). Positive attitude towards research has moderately significant positive correlations with verbal–linguistic intelligence, logical-mathematical intelligence, spatial intelligence, interpersonal intelligence and naturalistic intelligence subscales (r = 0.491; 0.375; 0.342; 0.352; 0.309, respectively) and slightly significant positive correlations with musical intelligence, bodily-kinesthetic and interpersonal intelligence subscales (r = 0.219; 0.168; 0.293, respectively) (p < 0.05). However, there is no significant correlation between the positive attitude towards researchers subscale and the musical intelligence subscale (p > 0.05). The positive attitude toward researchers subscale has significant positive moderate correlations with verbal–linguistic, interpersonal, and intrapersonal intelligence subscales (r = 0.512; 0.525; 0.475, respectively) and significant positive moderate correlations with logical-mathematical, bodily-kinesthetic, spatial, and naturalistic intelligence subscales (r = 0.190; 0.208; 0.206; 0.215, respectively) (p < 0.05). In addition, there are significant negative moderate correlations between the STEM Value-Expectation Evaluation Scale and the unwillingness to help researchers and negative attitudes toward researchers subscales (r = −0.356; r = −0.348, respectively) (p < 0.05). In addition, there are significant positive moderate correlations between STEM Value-Expectations Assessment Scale and positive attitudes towards researchers and its subscales (r = 0.446; r = 0.313, respectively) (p < 0.05). However, there is no significant correlation between STEM Value-Expectations Assessment Scale and musical intelligence subscale (p > 0.05). There are significant positive low correlations between STEM Value-Expectations Assessment Scale and bodily-kinesthetic, spatial and interpersonal intelligence subscales (r = 0.211, 0.254, 0.299, respectively) (p < 0.05). There are significant positive moderate correlations between the STEM Value-Expectation Assessment Scale and the verbal–linguistic, logical-mathematical, interpersonal and naturalistic intelligence subscales (r = 0.324; 0.328, 0.401, 0.354, respectively) (p < 0.05). In summary, there is no correlation between STEM values and expectations and musical intelligence. However, as verbal–linguistic, logical-mathematical, bodily-kinesthetic, spatial, interpersonal, interpersonal and naturalistic intelligence levels increase, STEM values and expectations also increase. In addition, as verbal–linguistic, interpersonal, interpersonal intelligence and naturalistic intelligence increase, reluctance to help researchers and negative attitudes towards research decrease, and positive attitudes towards research and researchers increase. As reluctance to help researchers and negative attitudes towards research increase, expectations towards STEM decrease. As positive attitudes towards research and researchers increase, expectations towards STEM also increase. According to the results, as the reluctance to help researchers and negative attitudes towards research increase, motivation towards STEM decreases. This finding shows that as the positive attitudes of female leadership candidates towards helping researchers and research increase, motivation towards STEM increases. In Table 7, the insignificant relationships presented regarding the mediating effect of attitude towards scientific research on the effect of multiple intelligences on STEM value expectation were removed from the model. The models examined within this scope are presented below.

The results of the mediator variable analysis for the models shown in Table 8 are presented in Table 9. In addition, the path diagrams for the mediator models are given in Figure 1. Model 1 shows the mediating effects of unwillingness to help researchers, negative attitudes toward research, positive attitudes toward research, and positive attitudes toward researchers on the relationship between verbal–linguistic intelligence and STEM value expectation evaluations. When all four mediator variables are included in the model, the path coefficients from unwillingness to help researchers, negative attitudes toward research, and positive attitudes toward researchers to STEM value expectation evaluations are not significant (p > 0.05). Therefore, unwillingness to help researchers, negative attitudes toward research, and positive attitudes toward researchers do not mediate the correlation between verbal–linguistic intelligence and STEM value expectation evaluations. Therefore, these variables are removed from the model. As a result, only the variable positive attitudes toward research remains as the mediator variable. The value RMSEA = 0.109 for this model indicates that there is no model-data fit. The values of χ²/df = 2.890, GFI = 0.921, NFI = 0.928, CFI = 0.951 indicate that there is an acceptable model fit. In the analysis, when there is no mediator variable, the standardized path coefficient from verbal–linguistic intelligence scores to STEM value expectation assessment scale scores is 0.32 and is significant (p < 0.05). In Model 1, where positive attitudes towards the research variable are included as a mediator variable, the standardized path coefficient from verbal–linguistic intelligence scores to positive attitudes towards research scores (β = 0.52) and the standardized path coefficient from positive attitudes towards research scores to STEM value expectation assessment scale scores (β = 0.41) are significant (p < 0.05). When the mediator variable is included in the model, the standardized path coefficient from verbal–linguistic intelligence to STEM value expectation assessments is 0.11 and is not significant (p > 0.05). After the mediator variable is included in the model, the predictor variable does not have a significant effect on the predicted variable, so the mediator variable has a full mediation effect. Therefore, the effect of verbal–linguistic intelligence on STEM value expectation for female health sciences faculty students is fully mediated by positive attitudes towards research. In addition, the indirect effect of verbal–linguistic intelligence scores on STEM value expectation assessment scale scores through positive attitudes towards research is 0.21. In other words, as the verbal–linguistic intelligence level increases for female health sciences faculty students, it increases STEM value expectations through the effect of positive attitudes towards research.

As seen in Table 8, Model 2 examines the mediating effect of positive attitudes toward research and positive attitudes toward researchers on the effect of logical-mathematical intelligence on STEM value expectation evaluations. When both mediating variables are included in the model, the path coefficients from positive attitudes toward researchers to STEM value expectation evaluations are not significant (p > 0.05). Therefore, positive attitudes toward researchers are removed from the model because there is no mediating effect on the effect of logical-mathematical intelligence on STEM value expectation evaluations. Finally, only positive attitudes toward the research variable remain as the mediating variable. As seen in Table 9, the model fit index for this model, RMSEA = 0.114, indicates that there is no model-data fit; while the values of χ²/df = 3.051, GFI = 0.917, NFI = 0.922, CFI = 0.946 indicate that there is an acceptable model fit. As a result of the analyses, the standardized path coefficient from logical-mathematical intelligence to value-expectation evaluations towards STEM is 0.33 and this effect is statistically significant (p < 0.05). In Model 2, where positive attitude towards research is included as a variable in the model, the standardized path coefficient from logical-mathematical intelligence dimension scores to positive attitude towards research dimension scores (β = 0.39) and the standardized path coefficient from positive attitude towards research dimension scores to value-expectation evaluations towards STEM scale scores (β = 0.40) are seen to be statistically significant (p < 0.05). When the mediator variable is included in the model, the standardized path coefficient from logical-mathematical intelligence to value-expectation evaluations towards STEM is 0.17 and is statistically significant (p < 0.05). The fact that the significant effect from the predictor variable to the predicted variable continues after the mediator variable is included in the model indicates that the mediator variable has a partial mediator effect. Accordingly, it can be said that the positive attitude towards the research variable has a partial mediating effect on the effect from logical-mathematical intelligence to STEM value expectation for women’s health sciences faculty students. In addition, the indirect effect from logical-mathematical intelligence dimension scores to STEM-focused value expectation assessment scale scores via positive attitude towards the research variable was determined as 0.16. In other words, the increase in the logical-mathematical intelligence level of women’s health sciences faculty students increases their value expectations towards STEM with the effect of positive attitude towards research. As seen in Table 8, Model 3 examines the mediating effect of positive attitudes towards research and positive attitudes towards researchers on the effect of bodily-kinesthetic intelligence on STEM value expectation evaluations. When both mediating variables are included in the model, the path coefficients from bodily-kinesthetic intelligence to positive attitudes towards research and from positive attitudes towards researchers to STEM value expectation evaluations are not significant (p > 0.05). This suggests that positive attitudes toward research or positive attitudes toward researchers do not have a mediating effect on the relationship between bodily-kinesthetic intelligence and STEM value expectancy assessments.

As seen in Table 8, Model 4 examines the mediating effect of positive attitudes towards research and positive attitudes towards researchers on the effect of spatial intelligence on STEM value expectation evaluations. When both mediation variables are included in the model, the path coefficients from positive attitudes towards researchers to STEM value expectation evaluations are not significant (p > 0.05). Therefore, positive attitudes towards researchers do not have a mediating effect on the effect of spatial intelligence on STEM value expectation evaluations and therefore it was removed from the model. As a result, only positive attitudes towards the research variable remained as a mediating variable. As seen in Table 9, the model fit index for this model, RMSEA = 0.110, indicates that there is no model-data fit; while the values of χ²/df = 2.900, GFI = 0.920, NFI = 0.925, CFI = 0.949 indicate that there is an acceptable model fit. As a result of the examination, it is seen that the standardized path coefficient from spatial intelligence to value-expectation evaluations towards STEM is 0.25 and this effect is statistically significant (p < 0.05). Positive attitude towards research is included as a variable in the model. In Model 2, the standardized path coefficient from spatial intelligence dimension scores to positive attitude towards research dimension scores (β = 0.36) and the standardized path coefficient from positive attitude towards research dimension scores to value-expectation evaluations scale scores for STEM (β = 0.43) are statistically significant (p < 0.05). When the mediator variable is included in the model, the standardized path coefficient from spatial intelligence to value-expectation evaluations towards STEM is 0.10 and is not statistically significant (p > 0.05). The fact that the significant effect from the predictor variable to the predicted variable does not continue after the inclusion of the mediator variable in the model shows that the mediator variable has a full mediator effect. Accordingly, it is seen that the positive attitude towards the research variable has a full mediation effect on the effect from spatial intelligence to the value-expectation towards STEM for female health sciences faculty students. In addition, the indirect effect from the spatial intelligence dimension scores to the value-expectation assessment scale scores for STEM through the positive attitude towards research variable was determined as 0.16. In other words, the increase in the spatial intelligence level of female health sciences faculty students increases their value-expectation towards STEM with the effect of the positive attitude towards research.

In Model 5, the mediating effect of unwillingness to help researchers, negative attitude towards research, positive attitude towards research and positive attitude towards researchers on the effect of interpersonal intelligence on value-expectation evaluations towards STEM is examined. When four mediator variables are included in the model, the path coefficients from the variables of unwillingness to help researchers, negative attitude towards research and positive attitude towards researchers to value-expectation evaluations towards STEM are not statistically significant (p > 0.05). In this case, the variables of unwillingness to help researchers, negative attitude towards research and positive attitude towards researchers do not have a mediating effect on the effect of interpersonal intelligence on value-expectation evaluations towards STEM and are removed from the model. In the last case, only the variable of positive attitude towards research is included as a mediator variable. As seen in Table 9, the value of RMSEA = 0.116 among the model fit indices for this model shows that there is no model data fit; χ²/df = 3.134, GFI = 0.913, NFI = 0.920, CFI = 0.943 values indicate that there is an acceptable model fit. As a result of the examination, the standardized path coefficient from interpersonal intelligence to value-expectation evaluations for STEM is 0.30 and this effect is statistically significant (p < 0.05). When positive attitude towards research is included in the model as a variable, the standardized path coefficient from interpersonal intelligence dimension scores to positive attitude towards research dimension scores in Model 5 is (β = 0.30), and the standardized path coefficient from research attitude dimension scores for STEM to value-expectation evaluation scale scores is (β = 0.42) and this effect is statistically significant (p < 0.05). When the mediator variable is included in the model, the standardized path coefficient from interpersonal intelligence to value-expectation evaluations for STEM is 0.18 and this effect is statistically significant (p < 0.05). After the mediator variable is included in the model, the significant effect from the predictor variable to the predicted variable continues, indicating that the mediator variable has a partial mediator effect. Accordingly, it can be said that the positive attitude towards the research variable has a partial mediator effect in the effect of interpersonal intelligence on the value-expectation towards STEM for female health sciences faculty students. In addition, it is determined that the indirect effect of the interpersonal intelligence dimension scores on the STEM-focused value-expectation assessment scale scores through the positive attitude towards the research variable is 0.12. In other words, the increase in the interpersonal intelligence level of female health sciences faculty students increases their value-expectations towards STEM with the effect of the positive attitude towards research.

In Model 6, the variables of unwillingness to help researchers, negative attitude towards research, positive attitude towards research and positive attitude towards researchers are found to mediate the effect of interpersonal intelligence on value-expectation evaluations towards STEM. When four mediator variables are included in the model, the path coefficients from the variables of unwillingness to help researchers, negative attitude towards research and positive attitude towards researchers to value-expectation evaluations towards STEM are not statistically significant (p > 0.05). In this case, the variables of unwillingness to help researchers, negative attitude towards research and positive attitude towards researchers do not mediate the effect of interpersonal intelligence on value-expectation evaluations towards STEM and are removed from the model. In the last case, only the variable of positive attitude towards research is included as a mediator variable. As seen in Table 9, the value of RMSEA = 0.117 among the model fit indices for this model indicates that the model data do not fit; χ²/df = 3.168, GFI = 0.913, NFI = 0.920, CFI = 0.943 values indicate acceptable model fit. As a result of the examination, the standardized path coefficient from personal intelligence to value-expectation evaluations towards STEM is 0.40 and this effect is statistically significant (p < 0.05). In Model 6, where positive attitude towards research is included as a variable in the model, the standardized path coefficient from personal intelligence dimension scores to positive attitude towards research dimension scores (β = 0.36) and the standardized path coefficient from research attitude dimension scores for STEM to value-expectation evaluations scale scores (β = 0.37) are statistically significant (p < 0.05). When the mediator variable is included in the model, the standardized path coefficient from personal intelligence to value-expectation evaluations towards STEM is 0.27 and is statistically significant (p < 0.05). The fact that the significant effect from the predictor variable to the predicted variable continues after the mediator variable is included in the model shows that the mediator variable has a partial mediator effect. Accordingly, it can be said that the positive attitude towards the research variable has a partial mediator effect on the effect coming from personal intelligence on the value expectation towards STEM for women health sciences faculty students. In addition, it is determined that the indirect effect from the personal intelligence dimension scores to the STEM-focused value expectation assessment scale scores through the positive attitude towards the research variable is 0.13. In other words, the increase in the personal intelligence level of women health sciences faculty students increases their value expectations towards STEM with the effect of the positive attitude towards research.

In Model 7, the mediating effects of the variables of unwillingness to help researchers, negative attitude towards research, positive attitude towards research and positive attitude towards researchers on the effect of natural intelligence on value-expectation evaluations towards STEM are analyzed. In the case of including four mediator variables in the model, the path coefficients from the variables of negative attitude towards diseases and positive attitude towards diseases to value-expectation evaluations towards STEM are significant (p > 0.05). In this case, the effect of helping with health improvements, negative attitude towards conditions and positive attitudes towards improvements on the effect of natural intelligence on value-expectation evaluations towards STEM is not specified and is removed from the model. In the last case, only the positive attitude variable is included as a mediator variable. Table 9 shows that there is no model-data fit with the model fit index RMSEA = 0.107 for modifying this model. The values of χ²/df = 2.826, GFI = 0.922, NFI = 0.927, CFI = 0.951 indicate an acceptable model fit. As a result of the examination, when there is no mediator variable, the standardized path from naturalistic intelligence to value-expectation evaluations towards STEM is 0.35 and this effect is consistently significant (p < 0.05). In Model 7, where positive attitudes towards research are included as model variance, the standardized paths from naturalistic breadth dimension scores to positive attitudes towards customers dimension scores (β = 0.32) and the standardized path from positive attitudes towards standards dimension scores to range scores of value-expectation evaluations towards STEM are consistently significant overall (p < 0.05) (β = 0.39). The change in the standardized path from naturalistic intelligence to value-expectation evaluations towards STEM is 0.23 and is permanently significant (p < 0.05). The fact that the significant effect from the predictor variable to the predicted variable continues after the inclusion of the mediator variable into the model shows that the mediator variable has a partial mediator effect. Accordingly, it can be said that the positive attitude towards research variable has a partial mediator effect in the effect of naturalistic intelligence on value-expectations towards STEM among female health sciences faculty students. In addition, the indirect effect of the positive attitude towards research variable from the naturalistic intelligence dimension scores to the value-expectation evaluation scale scores towards STEM is determined as 0.12. In other words, the increase in the intrapersonal intelligence levels of female health sciences faculty students increases their value-expectations towards STEM with the effect of the positive attitude towards research.

In conclusion, musical intelligence does not affect STEM values and expectations. An increase in bodily-kinesthetic intelligence enhances STEM values and expectations; however, no mediating effect was detected on attitudes towards scientific research. An increase in linguistic, logical-mathematical, spatial, interpersonal, intrapersonal, and naturalist intelligence increases STEM values and expectations, with only the positive attitude towards research acting as a mediating role in this relationship. In other words, high scores in verbal–linguistic intelligence, logical-mathematical intelligence, spatial intelligence, interpersonal intelligence, intrapersonal intelligence, and naturalistic intelligence lead to positive attitudes toward research. This positively affects female students’ STEM values and expectations.

This study found that most female leadership candidates had low-moderate reluctance to help researchers and negative attitudes toward research, moderate positive attitudes toward research, and high positive attitudes toward researchers. It suggests that female leadership candidates who are likely to take part in organisational development in the future will exhibit high level positive attitudes towards research. Çelik et al. (2015) investigated nursing students’ anxiety levels and attitudes towards scientific research and various variables affecting them. As a result, it was found that the anxiety level of the students towards doing research was not very high and their attitudes towards doing research were positive. Björkström et al. (2003) investigated Swedish undergraduate nursing students’ attitudes towards and awareness of research and development in nursing. They focused on 201 students. As a result, they stated that nursing students’ attitudes towards research were positive. Kes and Öztürk-Şahin (2019) worked with nursing students. They worked with 162 students to determine students’ anxiety and attitudes towards scientific research. According to the results of the study, it was found that students’ attitudes towards research were positive. Korkmaz et al. (2011) developed an attitude scale to determine pre-service teachers’ attitudes towards scientific research. There were a total of 1,085 students in the study group. According to the results of the study, it was found that pre-service teachers’ attitudes towards research were positive. Ercoşkun (2019) examined pre-service teachers’ attitudes towards scientific research with various variables. In the study, 381 students were studied. Attitudes Towards Scientific Research Scale (ATSRS) was used as a data collection tool. According to the results of the study, it was found that pre-service teachers’ attitudes towards research were positive. Çakmak et al. (2015) examined pre-service social studies teachers’ attitudes towards scientific research. They worked with 259 students in the study in which the survey model was used. According to the results of the study, it was found that pre-service social studies teachers’ attitudes towards research were positive. Yılmaz et al. (2020) studied 57 undergraduate students to determine their attitudes towards scientific research. As a result, it was found that the students’ attitudes towards scientific research were positive. Halabi (2016) investigated the attitudes of nursing students in Saudi Arabia towards research. The study was conducted with 244 final year students in a three-campus nursing college in three regions of Saudi Arabia. As a result of the study using the survey model, it was found that the students’ attitudes towards scientific research were positive. Halabi and Hamdan-Monsour (2010) evaluated Jordanian nursing students’ attitudes towards nursing research. The study included 612 senior nursing undergraduate students studying in Oman, Jordan. As a result of the research using the survey model, it was determined that nursing students had positive attitudes towards research. Studies in the field overlap with the finding in this study. It was also found that different results were obtained in different studies. Polat (2014) aimed to examine the attitudes of education faculty students towards scientific research. In the research conducted with 417 students in total, a survey model was established. Attitudes Towards Scientific Research Scale (ATSRS) and a questionnaire were used as data collection tools. At the end of the study, it was seen that students’ attitudes towards scientific research were at a moderate level. Biçer et al. (2013) investigated the willingness of prospective Turkish teachers to scientific research. The study focused on 312 prospective teachers studying at two different state universities. At the end of the research, it was seen that the students’ attitudes towards scientific research were at a moderate level. A lack of self-belief in terms of mathematics and science aptitude limits girls’ and women’s science aspirations much more than their performance. A survey of more than 2,000 girls aged 15 to 19 in the Asia-Pacific region suggested that only 12% continued to study science subjects even though more than 50% were considering them when they were younger. It suggested that female students made these decisions due to gender discrimination, the difficulty of the courses, and their perception that they received inadequate social support from their families and teachers. Girls are more likely to experience math anxiety than boys. Mothers are also likely to be more anxious than fathers. They may be more likely to transmit math anxiety to their children, especially girls (UNESCO, 2024). These findings are consistent with the findings of this study. There are a number of challenges to women’s advancement in academic leadership. Some of these are unconscious biases and deep-rooted gendered expectations against women. Cultural beliefs about gender and gender biases that stem from workplace structures, practices, and interaction patterns that unintentionally favor men. Unconscious gender bias is a powerful but often invisible barrier to women’s advancement. Men have characteristics that are socially and anthropologically more adaptable to leadership roles, while women have poorer self-images, greater lack of confidence in women than men, and less career orientation and less interest in pursuing a career (Ayyildiz and Banoglu, 2024; Gamero, 2024; Marques et al., 2024).

It can be said that the verbal–linguistic, mathematical-logical, musical, bodily-kinesthetic, spatial-kinesthetic, spatial, naturalistic intelligence levels of most of the female students studying at the Faculty of Health Sciences are at medium level, while their interpersonal intelligence and intrapersonal intelligence levels are medium-high. Studies supporting the findings of these studies are frequently found in the literature. Mathematical ability and knowledge are critical for developing STEM skills and working in STEM fields. However, girls’ confidence in these subjects tends to be lower than boys’, even when they perform well. It is estimated that one in five people feel anxious about maths, but anxiety levels are higher among girls. In all educational systems participating in TIMSS in 2019, except Bahrain and Egypt, male students were found to be significantly more confident in mathematics than female students (UNESCO, 2024). One factor that has been shown to negatively affect mathematics performance and acquisition is mathematics anxiety. Math anxiety is generally more pronounced in women than in men. Gender stereotypes are believed to play a role in the high levels of mathematics anxiety (MA) reported by female students. Justicia-Galiano et al. (2023) examined gender stereotypes about mathematics anxiety (MA), its ability and affective components. It was investigated whether gender differences in MA were associated with gender stereotype beliefs in both ability and affective aspects. A total of 257 secondary school students completed measures of mathematics-related and gender stereotypes. It was examined whether there was a stereotypical belief that male and female adolescents experience mathematics anxiety to different degrees. The results clearly show that both gender groups believed that female students experience higher levels of mathematics anxiety than male students. The results regarding beliefs about mathematics anxiety are consistent with the pattern of gender differences found for beliefs about anxiety in general. Both male and female students believe that girls are more likely than boys to experience general anxiety-related emotions such as tension and worry. These findings suggest that negative emotions such as distress, sadness, fear, shame, and guilt in mathematics performance are experienced more by women. Regarding the emotional aspect of math anxiety experienced by both boys and girls, it was found that girls believed that they experienced math anxiety more than boys. The results showed that women tended to experience more negative emotions than men. It was concluded that gender differences in math anxiety may reflect actual levels of math anxiety rather than gender bias. Regarding the math ability stereotype, both girls and boys held egalitarian or pro-female views. It could be argued that being aware of the stereotype that boys will perform better than girls may have negative consequences for some girls, as the perception of these ideas may lead them to feel insecure and anxious in math situations. This finding shows that the traditional stereotype that positively discriminates against boys is gradually decreasing in Western societies. Girls who held the traditional math ability stereotype exhibited lower math self-confidence, lower math performance, and higher math anxiety. This finding highlights the relationship between gender, mathematical ability, self-confidence, beliefs about success, and mathematical anxiety. This suggests that female students who are likely to be future leaders should be more motivated to be successful in mathematical fields. Developing girls’ math skills is crucial to STEM. Van Mier et al. (2019) assessed the mathematics anxiety and mathematics performance of 124 s and fourth grade children, 67 girls and 57 boys. Although boys and girls showed roughly equal levels of mathematics anxiety and performed similarly on the arithmetic task, correlation analyses revealed that only in girls did mathematics anxiety correlate significantly with mathematics performance. Higher levels of math anxiety only significantly and negatively moderated math performance in girls. The results showed that math anxiety is already negatively linked to math performance in girls as early as second grade. Math anxiety has lifelong consequences on girls’ math performance, causing them to later avoid STEM careers. Preventing the development of math anxiety, especially in girls, from early childhood is essential. For girls who have encountered the negative consequences of math anxiety on math performance it is imperative to reduce the level of math anxiety. Therefore, intervention studies should be conducted to reduce girls’ mathematics anxiety. Gender-related effects of intervention programs implemented from an early age should be taken into account. İzci and Sucu (2014) examined the multiple intelligence profiles of university students. A total of 722 undergraduate students studying at a state university were included in the study. In the study, the intelligence areas of university students were evaluated with the “Multiple Intelligences Questionnaire.” Significant differences were found between the intelligence areas of university students in terms of gender variable. It was observed that female students had higher mean scores in visual–spatial and musical-rhythmic intelligence than male students. Pehlivan (2008) worked with 934 students to examine the differences between intelligence areas according to gender. The Multiple Intelligences Inventory was used to determine multiple intelligence areas. It was found that female students were more developed in verbal–linguistic, visual–spatial, musical-rhythmic and self-directed intelligence areas than male students. Loori (2005) examined the differences in the intelligence preferences of male and female students learning English as a second language in higher education institutions in the United States. In the study, it was found that female students’ interpersonal intelligence and intrapersonal intelligence levels were medium-high. İzci et al. (2007) examined the scores of the students enrolled in the dershanes in various intelligence areas in terms of their gender. A total of 387 students were included in the study and the Multiple Intelligences Scale was used. When the multiple intelligence areas of female students were examined, it was observed that they had high levels of linguistic, physical, musical, social and introverted intelligence areas. Altınok (2008) determined the multiple intelligence areas of students studying at the school of physical education and sports according to various variables. It was found that women’s musical-rhythmic intelligence levels were higher than men’s musical-rhythmic intelligence levels. In addition, it was determined that women’s visual–spatial intelligence and bodily-kinesthetic intelligence levels were higher than men’s visual–spatial intelligence and bodily-kinesthetic intelligence levels. These findings are similar to the findings of the study. Participating in physical exercises protects young people’s mental health and increases their motivation. Exercises can be said to improve mood and cognitive function. It reduces depressive symptoms, increases self-efficacy and improves emotional regulation (Li Z. et al., 2024). In addition, physical activities have positive effects on health motivation, appearance motivation, leisure motivation, ability motivation and social motivation of young people (Li J. et al., 2024). Teachers play a critical role in learning activities using multiple intelligence systems. It has been found that when teachers motivate their students to arts and sports, students’ motivation to study is also high. Making music has also been found to increase motivation (Immerz et al., 2024). This suggests that participation in physical and artistic activities would be beneficial for the motivation of female leadership candidates.

When the value expectations of future female leadership candidates for the STEM field were evaluated, it was found that 81.13% of the students had a medium level of value expectation evaluation level for the STEM field. This situation suggests that female students are not positively motivated for STEM in the family, school and peer environment. It is possible to say that males are less motivated in social sciences and females are less motivated in math-oriented fields such as physical sciences, technology, engineering and mathematics. It was determined that girls had a negative self-perception especially in mathematics. Therefore, it negatively affects girls’ motivation towards STEM. Female students negatively evaluate their skills, competencies and potential in mathematics. The negative motivation of female students causes them to move away from STEM fields in their career expectations and plans (Lazarides and Lauermann, 2019; Lazarides et al., 2016). Dietrich and Lazarides (2019) examined the development of motivational belief patterns in mathematics by gender over a school year and career plans in mathematics-related fields. They found that girls’ STEM motivation was not at a high level. This finding is in line with the research finding. This suggests that female leadership candidates who will take their place in organizational development need more motivation in STEM. The study showed that musical intelligence does not affect STEM values and expectations. A measurement tool measuring STEM values and expectations was used in this study. A measurement tool regarding STEAM motivation was not used. STEM is based on technical and cognitive skills such as science, technology, engineering and mathematics. STEM education programs focus only on the technical aspects of their disciplines. STEAM (STEM+Arts) is based on science, technology, engineering, mathematics and art-creativity. STEAM Education is a learning approach that uses Science, Technology, Engineering, Art and Mathematics as access points to guide students’ questioning, dialogue and critical thinking. Musical tendencies are related to art. The instrument measuring STEM motivation used in this study did not predict musical intelligence and art. In this case, it is thought that musical intelligence does not predict STEM motivation (Huang, 2020; Özer and Demirbatır, 2023; Turhal, 2020).

It is determined that most female leadership candidates have a low-medium level of reluctance to help researchers. It is seen that their positive attitudes towards research are medium and their positive attitudes towards researchers are high. Most of them have medium levels of verbal–linguistic, mathematical-logical, musical, bodily-kinesthetic, spatial and naturalistic intelligence. It is determined that their interpersonal intelligence and interpersonal intelligence levels are medium-high. 81.13% of female students in the Faculty of Health Sciences have a medium level of value-expectation evaluation level towards STEM. It can be said that it is possible to improve women’s skills towards STEM. Recommendations for parents, teachers, researchers, policy makers and female students are presented below for future women leaders to take a strong role in organisational development.

Suggestions for families;

Suggestions for teachers;

Suggestions for researchers.

Recommendations to policy makers;

Recommendations for female leadership candidates;