Demographics
After the data were collected and cleaned, the sample size was 641 women. The respondents were demographically diverse (Table 1), with most reporting Age between 27-42 years (62.03%), Caucasian race (76.67%), residing Outside the U.S. (30.23%) or in the U.S. West (24.84%). The most common living situations were Urban (48.94%) with a household size of 2-3 people (64.01%). Most respondents reported having 9-15 Years in STEM (38.92%), a Doctorate Level STEM degree (41.34%), and a current occupation of Life Scientist (30.48%) or Computer Scientist (26.77%), as shown in Table 2. Most of the respondents reported that they did participate in STEM-specific social media (n = 405, 66.28%).
Autonomy, Competence, and Relatedness Scores
This study resulted in several key findings that underscore the importance of autonomy, competence, and relatedness as fundamental factors influencing the persistence of women in STEM. The sample demonstrated high scores (Table 3) for Autonomy Satisfaction (M = 15.75, SD = 3.84), Competence Satisfaction (M = 16.70, SD = 4.20), and Relatedness Satisfaction (M = 16.69, SD = 3.65).
The group produced lower scores (Table 4) for Autonomy Frustration (M = 12.07, SD = 4.20), Competence Frustration (M = 10.17, SD = 4.52), Relatedness Frustration (M = 8.82, SD = 4.11). ANOVA tests (Table 5) showed relationships between autonomy, competence, relatedness, and persistence. The Combined Satisfaction of autonomy, competence, and relatedness was significantly different between levels of Years in STEM (F(3, 473) = 3.19, p = .023), with the mean for 6-8 years (M = 46.41, SD = 10.34) significantly smaller than for 30+ years (M = 51.05, SD = 9.73, p = .046). Also, Competence Satisfaction (F(3, 465)= 11.90, p < .001) was significantly different between levels of Years in STEM. The mean scores for Competence Satisfaction were significantly lower for women in less-experienced groups than for those with the most Years in STEM.
ANOVA tests (based on an alpha value of .05) also showed significant differences in Combined Frustration (F(3, 467) = 3.69, p = .012) of autonomy, competence, and relatedness between levels of persistence (Table 5). Further analyses revealed significantly higher (p = .031) Combined Frustration for women with 6-8 Years (M = 32.04, SD = 11.12) and 9-15 Years (M = 31.21, SD = 10.02) than for those with 30+ Years (M = 26.11, SD = 8.75). Furthermore, Competence Frustration (F(3, 461) = 14.98, p < .001) was significantly different between levels of Years in STEM. The mean scores for Competence Frustration for women with 6-8 Years (M = 11.49, SD = 4.63, p < .001), 9-15 Years (M = 10.66, SD = 4.45, p < .001), and 16-29 Years (M = 9.18, SD = 4.11, p = .008) were all significantly greater than for those with 30+ Years (M = 6.56, SD = 3.13) in STEM. Competence Frustration for the group with 6-8 years in STEM was significantly higher than for those with 16-29 years (p < .001). Additionally, Competence Frustration for those with 9-15 years in STEM was significantly higher than for those with 16-29 years (p = .018) and those with 30+ years (p < .001). The effect sizes (eta squared) of the associations were small, indicating that other variables were involved. These findings provide statistical evidence supporting the relevance of SDT in understanding and improving women’s persistence in STEM.
Narrative Themes
The qualitative data complemented and enriched the quantitative findings by providing an in-depth look into the lived experiences of women in STEM. The open-ended questions garnered nearly 2300 narrative responses. The respondents described their experiences with persistence in seven Autonomy Satisfaction themes (Enjoyment, Choice and Input, Genuine Interest, Work-life Balance, Self-care and Health, Purpose, and Be Yourself), five Competence Satisfaction themes (Self-efficacy, Growth Mindset, Grit and Perseverance, Science Identity, and Effort), and five Relatedness Satisfaction themes (Social Support, Communal Benefit, Belonging, Role Models, and Communication Skills), all shown in Table 6. They also indicated five Autonomy Frustration themes (Disappointment, Lack of Choice, Lack of Work-life Balance, Lack of Enjoyment, Lack of Interest), three Competence Frustration themes (Lack of Knowledge, Lack of Career Benefits, Lack of Confidence), and two Relatedness Frustration themes (Lack of Relatedness, Lack of Role Models), shown in Table 7. Finally, there were six sociocultural themes (Discrimination and Bias, Career and Money, Problems with Men, Concern About Family, Academia Culture, and Problems with Women; Table 8). All of the themes derived from the qualitative data are presented below, with the frequency of code assignments, theme descriptions, and illustrative quotations. The narratives shared by participants provided valuable insights into how they navigate challenges and find support within their fields. These themes provide additional context for the quantitative scores obtained from the BPNSFS.
The thematic results enhanced the quantitative results, affirming the influence of autonomy, competence, and relatedness on persistence, but also highlighting the widely varied experiences of women in STEM. The integration of quantitative and qualitative results allowed for strong conclusions to be developed. The quantitative data provided statistical evidence of the positive relationship between need satisfaction and persistence in STEM, although the effect sizes of the theoretical constructs and sociocultural variables on persistence were small. This indicates that other influences must account for most of the observed associations. Given the diversity within the sample, which included participants from around the world, members of more than seven racial groups, spanning more than four decades in age and experience, representing more than eight STEM disciplines, at various stages of family life and identity intersections, it is not surprising that the results indicate additional variables at play. However, the totality of the findings emphasizes the need for policies and practices that foster the satisfaction of autonomy, competence, and relatedness to promote the persistence of women in STEM.
Integration of Quantitative and Qualitative Findings to Answer Research Questions
The demographic description, quantitative findings from the BPNSFS, and thematic results of open-ended responses provided a rich depiction of their experiences with persistence in STEM. The results from the data sets were aligned and integrated to form a more complete picture of the findings. Key quantitative and qualitative findings are presented together in Table 9 and are applied to answer the research questions.
RQ1: What Are the Levels of Autonomy, Competence, and Relatedness of Women Who Have Persisted in STEM?
The respondents, all of whom were STEM persisters, generally displayed high levels of satisfaction and low levels of frustration across the constructs of autonomy, competence, and relatedness. Persistence, measured as Years in STEM, was found to be associated with combined psychological need satisfaction (F = 3.19, p = .023) and inversely associated with combined psychological need frustration (F = 3.69, p = .012). A significant association was found between the individual construct of competence and persistence, suggesting that competence satisfaction (F = 11.90, p < .001) and competence frustration (F = 14.98, p < .001) were important to the participants’ persistence. The statistically significant associations between Years in STEM, need satisfaction, and need frustration, were small, suggesting that other variables also affected persistence.
RQ2: What Experiences and Influences Contribute to Women’s Persistence in STEM?
The respondents shared many influences on their motivations to persist, including supports and obstacles. The most frequently mentioned supportive influences included satisfaction of autonomy, competence, and relatedness. Autonomy satisfaction was supported by feelings of Enjoyment, Choice and Input, having Genuine Interest in STEM, maintaining Work-life Balance, having a sense of Purpose, prioritizing Self-care and Health, and feeling able to Be Yourself. Competence satisfaction was supported by experiencing strong Self-efficacy, having a Growth Mindset, demonstrating Grit and Perseverance, a sense of Science Identity, and making Effort. Relatedness satisfaction was experienced with Social Support, working for Communal Benefit, having a sense of Belonging, the presence of Role Models, and good Communication Skills. In addition to the satisfaction of psychological needs, women were also motivated to persist in STEM by factors related to Career and Money.
RQ3: How Do Women Explain Various Sociocultural Influences on Their Persistence in STEM?
Respondents reported numerous sociocultural influences on their persistence in STEM. Several sociocultural demographic factors, in addition to theoretical constructs, were found to be significantly associated with persistence. Quantitative results indicated significant associations between persistence and Age, Race, Area Type, Highest STEM Degree, Current Occupation, and Household Size. Within those variables, certain subgroups were identified as being significantly different in terms of persistence. For example, Age (large effect) and Household Size (small effect) were positively correlated with Years in STEM, meaning that participants with higher levels of persistence tended to be older and live in larger household groups. Within the Race variable, members of Caucasian and Other groups had higher persistence than Asian respondents. Respondents living in Rural areas demonstrated higher levels of persistence than those in Urban areas. Respondents with Doctoral Level degrees demonstrated significantly greater persistence than those with only Master’s Level and Bachelor’s Level degrees. Finally, those whose Current Occupation was STEM Educator had significantly higher persistence than Computer Scientists, suggesting that one’s occupation may affect her persistence.
From the narrative data, the most frequently mentioned sociocultural influences involved Discrimination and Bias. There were multiple types of Discrimination and Bias experienced, including gender discrimination, sexual harassment, and racism. Women who are members of multiple minoritized groups were perceived by others to experience greater levels of Discrimination and Bias than those in majority populations. Other sociocultural influences included considerations of Career and Money, interpersonal Problems with Men and Problems with Women, feeling Concern about Family and navigating Academia Culture. Most respondents described their experiences as complicated and challenging, but said they were still motivated to persist.
RQ4: How Do Women’s Experiences Align with Their Levels of Autonomy, Competence, and Relatedness?
There was strong alignment between quantitative levels of autonomy, competence, and relatedness and women’s experiences persisting in STEM. The quantitative results showed relatively high satisfaction and low frustration of autonomy, competence, and relatedness. Study participants shared thousands of personal examples, describing how they experience satisfaction and frustration of their basic psychological needs. The narrative evidence provided rich context and real-world examples of women’s experiences of autonomy, competence, and relatedness. Table 9 highlights results from both the quantitative and qualitative analyses side-by-side to show this alignment.
Summary of Findings
Women who persist in STEM tended to display high satisfaction of their basic psychological needs. Significant associations were found between Years in STEM, Combined Satisfaction (autonomy, competence, and relatedness) and Competence Satisfaction. Women also tended to perceive low levels of autonomy, competence, and relatedness frustration. Significant associations were found between Years in STEM, Combined Frustration, and Competence Frustration. While significant associations were found between some variables and persistence, the effect sizes were small, indicating the presence of other influences. Sociocultural factors were described as contributing both positive and negative influences on women’s motivations to persist in STEM.
Study Conclusions
After comprehensive analyses of the data, the researcher drew four conclusions from the study. These conclusions are supported by the findings of either the quantitative or qualitative analyses or both. The conclusions are discussed below, along with explanations of how they fit with previous research.
Conclusion 1: Women Who Persist in STEM Are Highly Satisfied Overall
Women who persist in STEM reported high levels of satisfaction in autonomy, competence, and relatedness. These three factors contribute to a sense of motivation for women in STEM, thereby encouraging persistence. Considering the constructs separately, competence may be particularly essential to women’s persistence in STEM, as this factor was individually and significantly associated with persistence for women in this study.
Implications for Scholarship. This study affirms previous findings that persistence is associated with the satisfaction of autonomy, competence, and relatedness. Women in this study, who have demonstrated persistence in STEM, exhibit high quantitative scores for construct satisfaction. These findings confirm previous research showing that autonomy (Mlambo & Mabokela, 2017; Schmitt et al., 2021; Simon et al., 2015; VanAntwerp & Wilson, 2018), competence (Buse et al., 2013; Nauta et al., 1998; Schaefers et al., 1997), and relatedness (Bostwick & Weinberg, 2022; Canaan & Mouganie, 2021; Dennehy & Dasgupta, 2017; Espinosa, 2011; Herrmann et al., 2016) positively affect persistence.
In addition to the significant positive association between overall satisfaction and persistence, there was a significant positive association between Competence Satisfaction and persistence. This invites further investigation into the relationship between competence and persistence, and contrasts somewhat with previous research. Work by Dasgupta was essential in the conception of this study, including her finding of the primacy of belonging as the critical factor in the retention of women in STEM (Dasgupta & Stout, 2014; National Science Foundation, 2016). The findings of this dissertation do not refute those of Dasgupta, since the current study also shows that persisters experience high relatedness satisfaction. However, the findings do provide evidence of the importance of competence in women’s decisions to persist in STEM.
Conclusion 2: Women in STEM Persist Despite Their Negative Experiences with Discrimination and Bias
Despite facing Discrimination and Bias, women who persist in STEM were able to navigate and overcome these challenges. Women in the study perceived experiences with Discrimination and Bias as unavoidable. Respondents described encounters with sexual harassment, racism, age discrimination, and ablism, with instances of overt mistreatment creating peaks above the baseline of unbelonging and frequent microaggressions. However, some shared that negative experiences inspired them to develop greater Grit and Perseverance to remain in STEM. Some women were motivated to persist in STEM so they could help change the system for future generations, as described in the theme of Communal Benefit. Others said they were willing to endure difficult situations due to their love and Enjoyment of science. Many respondents credited their own Social Support networks and Role Models with helping them navigate obstacles. Finally, a group of respondents said they were determined to stay in STEM for Career and Money. These and other themes show that women employ numerous strategies to persist in STEM careers despite widespread Discrimination and Bias.
Implications for Scholarship. This conclusion supports previous research findings that women who persist in STEM need strategies to handle discrimination (Casad et al., 2021; Fisher et al., 2020; Schmader, 2023). Women in this study persisted through harassment and discrimination by adapting their behavior (Buse et al., 2013), relying on other women for advice and support (Espinosa, 2011; Shapiro & Sax, 2011), focusing on deep subject interest (Verdín, 2021), expertise (Cech et al., 2011), sense of identity (Millar et al., 2022), and working toward change (Casad et al., 2021; Kong et al., 2020).
Conclusion 3: To Ensure Women Persist in STEM, Organizations Must Consider and Adopt Policies and Practices for Supporting Individual Autonomy, Competence, Relatedness, and Financial Equity with Their Male Colleagues
The third conclusion emphasizes the need to implement policies and practices that support women’s autonomy, competence, relatedness, and workplace equity to maintain and improve persistence. This study provides statistical and qualitative evidence of multiple variables involved in women’s persistence in STEM. Women in this study are highly satisfied with their autonomy, competence, and relatedness, which correlate with their persistence. The women who participated in this study described specific experiences that contribute to the satisfaction of those needs, which resulted in 765 passages coded for autonomy, 453 passages coded for competence, and 873 for relatedness. These persistence-supportive factors have been discussed in detail in previous sections. In addition, women provided 357 passages describing situations that made them feel controlled, undervalued, unwelcome, and in some cases mistreated and harmed. Taken together, these findings can inform the cultivation of environments and programs that support women’s persistence and minimize obstacles.
Implications for Scholarship. The top three autonomy themes were Enjoyment, Choice and Input, and Genuine Interest, indicating that women desire opportunities to explore their interests, choose activities that they enjoy, and decide how to spend their time. Previous research found that these autonomy practices support persistence (Brubacher & Silinda, 2019; Maltese & Cooper, 2017; Patall et al., 2010, 2017; Talley & Martinez Ortiz, 2017). Respondents described their needs for competence satisfaction, including Self-efficacy, Growth Mindset, and Science Identity, which confirms previous research on the importance of self-efficacy (Charleston & Leon, 2016; Rittmayer & Beier, 2009; Zeldin & Pajares, 2000), science identity (Millar et al., 2022; Stout et al., 2011), and growth opportunities (Fuesting et al., 2019; Posselt, 2018; R. Binning et al., 2019). This study confirms previous research findings on the need for relatedness to support persistence, by providing social support and belonging (DuBow et al., 2017), role models (Herrmann et al., 2016), mentors (A. M. Petersen, 2014), and communal goals (J. Allen et al., 2021; Belanger et al., 2020; Henderson et al., 2022; Peacock et al., 2020). In addition to receiving the support of others, women who persist in STEM also value opportunities to serve others (J. Allen et al., 2021).
Environments that support women’s persistence extend to their perceptions of equitable treatment on campus and at work. Financial equity for women in STEM goes beyond the numbers on a paycheck. The importance of Career and Money highlights the role of external factors in motivation, and supports previous findings that income positively affects women’s motivation (Schmitt et al., 2021). This finding has a place within the SDT framework, which recognizes a continuum of motivation, ranging from amotivation to intrinsic motivation (Howard et al., 2017; Ryan & Connell, 1989). External rewards are accounted for in various degrees of extrinsic motivation and lessen as motivation becomes more intrinsically regulated. Some of the respondents said their need for financial security also factored into their willingness to endure unpleasant workplaces. Previous research found that a need for financial safety can affect motivation along with psychological needs (Chen, Van Assche, et al., 2015). Respondents in this study shared their desires for flexible job options, maternity leave, and access to dependent care. These practices span multiple motivational variables and therefore are likely to require multi-pronged solutions, such as interventions to improve recruitment, mentoring, and climate (Casad et al., 2021) and practices designed to address psychological needs (Slemp et al., 2021). Organizations and educational institutions with environments that support these needs can help women persist in STEM.
Conclusion 4: There are Career Trajectory Points Where Risk of Attrition is More Likely to Occur
The findings of this study demonstrated that there are significant associations between satisfaction, frustration, and persistence. Statistical analyses showed that women with the greatest number of years in STEM had significantly higher levels of combined satisfaction and significantly lower levels of combined frustration than those with the least experience. Similarly, women in lower age groups and with less educational attainment had fewer average years in STEM than older women and those with doctoral degrees. These findings suggest that women in earlier career stages (in younger age groups, with less educational attainment, having fewer years in STEM) may be at greater risk for attrition than women with greater satisfaction and persistence levels. Therefore, early career environments that contribute to low satisfaction and/or high frustration may expose “leaks” in the STEM pipeline, perpetuating the gender gap problem. Knowing that persisters encounter higher frustration in their earlier careers can help them anticipate and prepare for obstacles.
Implications for Scholarship. Previous research has found many possible contributing factors to attrition from STEM, beginning in childhood and increasing throughout women’s educational and career pathways. This study confirms previous findings that women in STEM experience attrition-contributing factors including biases favoring men (Amon, 2017; Andrus et al., 2018; Moss-Racusin et al., 2012, 2018), low self-efficacy (Orenstein & American Association of University Women, 1994), lack of mentors and role models (Amon, 2017), and a lack of belonging (Dasgupta & Stout, 2014; Good et al., 2012; Goris, 2020; Moore, 2020; Xu & Lastrapes, 2021).
Implications for Practice
This study provides additional evidence of the need for autonomy, competence, and relatedness, for women to persist in STEM. While research-backed practices for supporting each construct have been discussed previously, there is a potential strategy that might address all three. These needs could be addressed by bringing together girls and women to engage in STEM activities well before students make decisions about college majors and career fields. Since students begin to develop attitudes about gender-appropriate career roles by kindergarten (Ceci et al., 2014), interventions to support their exploration of STEM must begin early.
For example, science class visits from women in STEM, robotics clubs with women as coaches and leaders, and science-related social groups for girls and women in STEM, would expose girls to peer and adult role models, allow space for curiosity and hands-on exploration, and give STEM professionals additional opportunities for networking and community development. Some existing programs such as career days, classroom visits from STEM professionals, and STEM field trips, might be good starting places for such interactions. To be most effective, the activities should be planned inclusively, to represent the broad diversity of people and careers in STEM. Particular attention should be given to including members of historically excluded groups, since they tend to experience greater difficulty accessing such opportunities. For women in STEM jobs, organizations must assess and address weaknesses in their policies and practices to develop more supportive work environments.
In addition to childhood support for STEM education and exploration, science educators and employers need to be aware of potentially weak areas in the pipeline. During transitions between schools, after common weed-out courses, and during early career stages, there may be higher risk of attrition for women in STEM. Providing additional support for autonomy, competence, relatedness, and financial equity around these points may help retain women in STEM educational and career pathways. Interventions that reinforce women’s volitional control, self-efficacy, growth mindset, and social support may contribute to retention.
Study Limitations
After it was launched, the researcher discovered that the survey offered an incomplete range of choices for the Race category, with no option listed for Latina or Hispanic respondents. Although respondents were able to write in an unlisted option, this omission may have caused some potential participants to feel excluded or to exit the survey. Also, the use of an anonymous online survey made it impossible to follow-up with respondents to probe for additional information. Another limitation was noted by a potential participant in a Reddit group, who pointed out that software engineers and other information technology (IT) professionals often focus on developing skills through experience and certificates, rather than obtaining traditional undergraduate and graduate degrees. By requiring an undergraduate degree to participate in this study, some women in IT were excluded, despite having persisted in STEM for more than six years. Finally, the sample size was not large enough to draw generalizable inferences about the relationships between variables, although this was not a goal of the study.
Internal Study Validity
Throughout the study, the researcher implemented reflexive practices to ensure study validity. She used a research journal to record observations, concerns, and ideas to help minimize the influence of her biases (Creswell & Creswell, 2018). The BPNSFS tool had been previously validated, and it was used with permission and scored according to the instructions provided by its originators (Van der Kaap-Deeder et al., 2020). Subject matter experts were consulted to ensure that the open-ended questions were appropriate for the study. A survey construction expert was consulted to assist with the design and flow of the instrument. The study was approved by the IRB prior to engaging participants, and no confidentiality or anonymity concerns were reported. Before launching the survey, a pilot test was conducted and minor adjustments were made. The researcher used software to collect and analyze data, which allowed for a transparent, documented process. A peer reviewer checked the researcher’s statistical analyses and qualitative coding to make sure they were done reliably.
Recommendations for Scholarship and Further Research
The quantitative and qualitative results of this study reinforce the tenets of SDT, providing further evidence that basic needs satisfaction is associated with persistence, including for this sample of women in STEM. Many themes revealed in the narrative responses also fit rather neatly within the construct “buckets.” Although the focus of this study was on factors that support persistence, it also found a significant negative association between frustration and persistence, which invites additional research. In addition to the scholarship implications noted after each conclusion above, this study provides additional support for the application of SDT and the BPNSFS. This is the first known study of women’s persistence in STEM that combines the BPNSFS instrument with qualitative evidence to triangulate results using mixed methods. The Cronbach alpha values calculated in this study provide additional evidence of the reliability of the BPNSFS as a research tool.
In future studies, the researcher recommends segmenting the data from this sample into narrower groupings, such as by career stage, to determine whether autonomy, competence, and relatedness scores and narrative themes vary by persistence level. This might provide additional insight into what women experience at early, mid, and later career stages, and what types of support are most valuable to women at throughout their careers in STEM. She would also like to conduct a study to interview persisters and non-persisters to gain additional insights into their experiences in STEM. Another research thread would be to design and implement a program that brings together girls and professional women in STEM, to investigate how the experience affects their perceptions of autonomy, competence, and relatedness and persistence. Ideally, this would be a longitudinal study, to determine whether and how it affects STEM persistence over time.