Identifying Root Causes for Deficit Participation Among Underrepresented Populations in STEM Career Pathways and Coursework by Laura M. Wilcock A thesis submitted in partial fulfillment of the requirements for the degree of MASTER OF EDUCATION with an emphasis in CURRICULUM AND INSTRUCTION WEBER STATE UNIVERSITY Ogden, Utah 05/04/2023 Approved by: Katarina Pantic, Ph.D. Megan Hamilton Ph.D. Noel Alton D.Sc. 2 Abstract The root causes of deficit participation by women and BIPOC populations in certain STEM fields i.e., physics, computer science and mathematics are not well understood. The goal of this project was to identify possible root causes based on the perceptions of underrepresented populations while they are still within the K-12 systems. Students were selected from grades 812 from a diverse school district from the northern Wasatch front area of the United States. These students all come from traditionally underrepresented populations in the STEM fields. Results indicated four major influences or root causes on student perceptions of math, science, and coding (computer science). They indicated that teacher influence, perceived/apparent difficulty of certain subjects, student engagement, and familial pressure to perform highly all have high impact on student persistence in STEM subjects as well as influence students desire to persist in such coursework. The implications of this research show the lasting influence a teacher and continued stereotypes of certain subjects being difficult to obtain an understanding or be successful in. Teacher and family influence can be both positive and negative factors in student retention and perseverance in these STEM subjects. 3 Table of Contents Abstract ..................................................................................................................................... 2 Table of Contents ...................................................................................................................... 3 Table of Tables ......................................................................................................................... 5 Table of Figures ........................................................................................................................ 6 Problem Statement ................................................................................................................... 7 Literature Review ...................................................................................................................... 8 CTE Coursework and Interventions ......................................................................... 9 Mentorship ............................................................................................................. 11 Complicated Causes .............................................................................................. 11 Methods and Design ............................................................................................................... 14 Research Design.................................................................................................... 14 Setting .................................................................................................................... 14 Sample ................................................................................................................... 16 Recruitment ............................................................................................................ 18 Data Collection and Procedure .............................................................................. 19 Data Analysis ......................................................................................................... 21 Results .................................................................................................................................... 24 I don’t think I could code to save my life ................................................................. 25 She Actually, Like, Knew How to Explain Math ...................................................... 27 All we do is worksheets .......................................................................................... 29 4 Family Matters........................................................................................................ 29 Discussion ............................................................................................................................... 29 Prior Experience..................................................................................................... 30 Perception of Difficulty and Self Efficacy ................................................................ 33 The Willingness to Engage .................................................................................... 36 Conclusions ............................................................................................................................ 37 Limitations ............................................................................................................................... 38 Implications and Future Work ................................................................................................. 38 References .............................................................................................................................. 40 References .............................................................................................................................. 41 Appendix A: Interview Script and Protocol .............................................................................. 48 Interview Script:...................................................................................................... 48 Interview Protocol:.................................................................................................. 48 Appendix B: Demographic Survey .......................................................................................... 50 Demographic Survey:............................................................................................. 50 Appendix C: IRB approval ....................................................................................................... 51 Appendix D: Letter to parents ................................................................................................. 53 Appendix E: Items for object-based elicitation: ....................................................................... 55 5 Table of Tables Table 1 .......................................................................................................................... 15 Table 2 .......................................................................................................................... 15 Table 3 .......................................................................................................................... 17 Table 4 .......................................................................................................................... 23 6 Table of Figures Figure 1 ........................................................................................................................ 22 Figure 2 ........................................................................................................................ 24 7 Problem Statement Initiatives for the integration of engineering and technology into science and math education have been at the forefront of many educational endeavors over the past decade (e.g., Estrada et al., 2016; Walan, 2019; Sullivan & Bers, 2018; Master et al., 2017; Zweben & Bizot, 2016). Using project-based learning curricula, teacher training, professional development, and other resources (e.g., Sullivan & Bers, 2018; Master et al., 2017; Olszewski-Kubilius et al., 2017; Walan, 2019), science, technology, engineering, and math (STEM) education have become an essential part of K-12 instruction. One of the main goals of these initiatives has been to address the lack of diversity that is present in STEM fields, particularly those dealing with engineering, computer science, and physical sciences. Focus on broadening participation is important because lack of diversity within STEM fields leads to narrowed perspectives, limited voice, and missed opportunities for field growth (Accenture, 2020; Charlesworth & Banaji, 2019; Green et al., 2002; Krishnan, 2020). Diverse workplaces can provide financial and intellectual benefits (Charlesworth & Banaji, 2019). Gender and ethnic diversity within STEM fields and subjects are necessary to meet the rigorous demands of productivity and innovation (Charlesworth & Banaji, 2019). Diversity of gender, race, and/or ethnicity can enhance and propel scientific development to greater gains. On the other hand, by not having a diverse workforce, the scientific community may suffer innovative consequences. Though many varied initiatives and seemingly strategic efforts have been made to address the diversity gap and bring more women and BIPOC persons into the fields of engineering, computer science, and physical sciences (Cimpian et al., 2020), data tracking of these efforts have shown little change in the gap and has led researchers to wonder why it exists (Reinking & Martin, 2018). 8 To better address the gap, it is important to understand the root cause(s) of underrepresentation. In identifying the root cause(s), steps can be made to begin to close the gap. To that end, this study asks the following research question (RQ): RQ: What are the root causes of STEM disparities among K-12 students from the perspective of middle and high school students? Literature Review It has been well-documented and established that gender-identifying women and Black, Indigenous, and persons of color (BIPOC) are disproportionately outnumbered in careers and fields related to the hard sciences, mathematics, engineering, and technology (STEM) (Okrent & Burke, 2021). Despite initiatives by institutions of higher education as well as K-12 academics, there remains a persistent gap in gender, ethnic, and culturally diverse populations in STEM achieving degree programs (Briggs, 2017, Cimpian, 2020). This issue has become recognized by the US Federal Government inasmuch as federal agencies have spent $3.4 billion dollars to support STEM education in 2010 alone (Douglas & Strobel, 2015; Okrent & Burke, 2021). Gender and racial inequality in the workforce have been a pervasive and troubling issues facing many career fields and workforce communities. Gender and racial imbalances within physics fields alone show that 75% of participants are White men (Blue et al., 2018). In computer science, engineering, and physics there is a 4:1 ratio of men to women participants (Cimpian et al., 2020). According to a study by Accenture (2020), only 2.8% of FORTUNE Global CEOs identify as women. Workplace diversity has become a crucial element of the hiring and recruitment process for employers as workforce diversity improves and increases productivity and overall performance of the industry especially considering the increased globalization of the workforce (Green et al., 2002; Krishnan, 2020). Cultural and gender diversity in the workplace can lead to greater productivity and competitive advantages over less diverse companies (Green et al., 2002; Ozgen et al., 2017). Studies also show that workforce teams that are composed of diverse cultural and ethnic 9 backgrounds have less conflict and tend to be more creative (BCG, 2017; Foma, 2014). Other studies conducted on employees of diverse workplaces report that employees have higher job satisfaction and feel as though they experience a more positive experience in dealing with coworkers when they are part of diverse groups (Ahmed, 2019). Consequently, a lack of diversity can lead to increased internal conflict among employees of homogeneous groups and can also lead to the members of cultural or ethnic communities feeling more ostracized (Foma, 2014). Companies and scientific communities can no longer ignore the implications and importance of having a diverse workforce if they want to succeed in an increasingly competitive and globalized community. Failing to connect people of different cultures and backgrounds has a negative impact on productivity, innovation, and overall job satisfaction. These innovative workforce communities can bring about positive change in the world of academia, business, and science. Closing representation gaps in STEM fields, therefore, is imperative. CTE Coursework and Interventions To mitigate the present and persistent workforce disparity, initiatives have been implemented in the form of career and technical education (CTE) coursework offerings at the middle and high school levels and interventions in K-6 settings (Burnett et al., 2016). These STEM and CTE initiatives can look different depending on the school setting and availability of facilitators, as well as participant needs (e.g., Master et al., 2017; Sullivan & Bers, 2019; Walan, 2019). Many programs in place are completed as after-school enrichment activities or as pushin programs that occur during the day and are facilitated by someone other than the classroom teacher (e.g., Sullivan & Bers, 2019; Walan, 2019). Programs can differ greatly in content, but they still have similar results. For example, robotics using KIBO, a small block-based robotics program cultivated similar attitudes and interests among participants as did makerspaces that utilized drama and 3D printing (e.g., Sullivan & Bers, 2018; Walan, 2019). 10 The results of the implementation of robotics and coding interventions as well as mentorship at the elementary level have demonstrated a positive mindset shift among women participants, as well as showed a pronounced increase in participants' desire to pursue a career along the same pathway as their mentor (Master et al., 2017; Sullivan & Bers, 2018). In longterm studies, after-school enrichment programs were found to lead to increased performance and participation among racial or ethnic minorities and women in more advanced math and science classes in high school settings (Olszewski-Kubilius et al., 2017). Outside of direct instruction style interventions, makerspaces in schools can function in a variety of environments and provide unique enrichment opportunities for participants. Arts and drama programs are easily integrated into makerspace along with coding and robotics (e.g., Barton et al., 2017; Walan, 2019). The longevity of these and similar studies vary in length. Often, the outcomes for longer periods of STEM exploration are more desirable (Master et al., 2017; Olszewski-Kubilius et al., 2017; Walan, 2019), but not exclusively beneficial. For example, despite only experiencing a single lesson in robotics, girls aged six reported greater levels of interest and self-efficacy in computer programming and robotics than girls in control groups where the hands-on robotic experience was not implemented (Master et al., 2017). There is also no shortage of online resources dedicated to computer science and coding that require only a computer or a tablet. Scratch and Code Combat are web-based coding resources used by organizations to generate interest among girls and young women in K-12 (e.g., Fields et al., 2015; Hite & Taylor, 2021; Yücel & Rizvanoglu, 2019). The diversity of programming options as well as materials may help dissolve preconceived anxieties surrounding computer competencies among disadvantaged students (Hite & Taylor, 2021; Yücel & Rizvanoglu, 2019). The digital divide among socioeconomically disadvantaged groups and affluent groups would be addressed in a classroom setting by the incorporation of these types of programs into daily routines (Ball et al., 2019). 11 Mentorship Mentorship for these projects was also found to be key to their success and participants showed a dramatic interest in continuing to use technology to create solutions to projects as well as an overwhelming interest in becoming engineers like their mentors (Walan, 2019). Mentoring seems to have a dramatic impact on identifying young women’s attitudes and selfefficacy in STEM programs, regardless of the type or even age demographic. Female participants in programs where mentorship is rigorous, inclusive, and accessible often selfreport higher levels of competency and positive attitude in all STEM subjects (Clark et al., 2018; Stoeger et al., 2013; Walan, 2019). Female mentors act as role models and help to bridge many of the disconnects between gender identifying women/girls’ participation and the adverse stereotypes they may feel exist in STEM fields (Clark et al., 2018; Stoeger et al., 2013). Complicated Causes Despite numerous efforts, STEM intervention programs have been marginally successful in generating interest among women and BIPOC participants, as they seemingly fail to make any lasting impact later as students’ progress through college (Leu & Arbeit, 2020; OlszewskiKubilius et al., 2017). Identifying the root causes of these gaps then becomes imperative for understanding why the disparity exists and the impetus for establishing a culture of change in K12 and higher education. Several studies have taken this issue and addressed it with an indepth survey of populations in K-12 education as well as higher education (e.g., Ball et al., 2019; Blue et al., 2018; Charlesworth & Banaji, 2019; Khan & Rodrigues, 2017; Master et al., 2017; Orkent & Burke, 2021; Reinking & Martin, 2018). One possible identified causation of the disparity within STEM fields may be rooted in deeply held attitudes relating to gender stereotypes that exist in culture and family structures (Blazev et al., 2017, Blue et al., 2018; Charlesworth & Banaji, 2019). Gender stereotypes among young children are seemingly evident by the time they reach first grade (Master et al., 2017). According to Olszewski-Kubilius et al. (2017), the failure of CTE programs at the middle 12 and high school levels is rooted in the fact that students' beliefs are already hardwired and in place by that point in their education. An individual’s self-efficacy also seems to play a role in whether students persist in STEM fields and careers. Results from a 2017 study involving urban, high poverty, and predominantly African American schools showed that most students reported a high affinity for STEM subjects but lacked the desire to pursue careers in those subject fields (Ball et al., 2019). This may be a direct result of their beliefs regarding expectancy for success and self-efficacy in those fields (Ball et al., 2019). The same study also concluded that there is a disparaging digital divide between affluent and highly impacted schools when it comes to digital literacy and exposure to technology (Ball et al., 2019), which may, in fact, contribute to the decrease in students reported self-efficacy in STEM subjects and the desire to pursue careers or degrees in such fields. Student attitudes regarding self-efficacy and confidence in STEM fields may influence their pursuit of degrees and careers in those disciplines, while prominent and pervasive stereotypes about gender and race may influence their likelihood of pursuing and remaining in such programs and/or careers (Blazev et al., 2017; Blue et al., 2018; Charlesworth & Banaji, 2019; Khan & Rodrigues, 2017). Self-efficacy and gender role stereotypes seemingly emerge even in relatively young children, especially children who grow up in communities where societal diversity within STEMrelated careers lacks gender and racial diversity (Niepel et al., 2019). In fact, many of these disparities in representation are beginning to be linked with students’ reported self-confidence and self-concept in math, science, technology, and engineering fields (Charlesworth & Banaji, 2019; Kurtz-Costes et al., 2014; Niepel et al., 2019). Self-concept is shaped by the environment and the experiences an individual has within that environment through connections with other individuals, including family, friends, and educational figures (Niepel et al., 2019, 2019; Pantic, 2020). Underrepresented groups, 13 including women, have often been found to perceive their own ability, and their self-efficacy based on their observations of societal norms (Niepel et al., 2019). As young adolescents, students already report specific gender bias when it comes to the educational subject matter, specifically, that women are better suited for tasks that involve verbal communication skills (Kurtz-Costes et al., 2014). In early elementary school settings, both boys and girls implicitly associate math with boys more than they do with girls (Charlesworth & Banaji, 2019). Students exposed to academic stereotypes are found to be more likely to endorse such stereotypes as they grow and mature (Charlesworth & Banaji, 2019; Kurtz-Costes et al., 2014; Niepel et al., 2019). Further, in a recent meta-analysis involving the Draw-A-Scientist test, as much as 73% of children involved in the research drew the scientist as male (Miller et al., 2018). Older children also drew male scientists more often than their younger counterparts did, supporting the idea that gender role stereotypes develop as children age and do or do not experience diverse representation (Miller et al., 2018) or there is not enough awareness around some professions, such as computer science (Pantic et al., 2018). Stereotypes can strongly influence interest and value in STEM fields, leading to future self-exclusion from these pathways (KurtzCostes, 2008). Computer science, physics, and engineering also seem to have developed their own gender and racial stereotypes. Children as young as six were found to endorse gender interest stereotypes regarding engineering, indicating in a survey, that engineering is reserved for boys, rather than girls (Master et al., 2021). These stereotypes only seem to become more prevalent as children age and mature through school (Master et al., 2021). Since students’ academic choices are influenced by their beliefs and self-concepts, it is only logical that fewer women and underrepresented individuals would choose to pursue careers and concepts that they feel are not designed for them (Blue et al., 2018; Master et al., 2021). 14 Methods and Design Though prevalent stereotypes and self-efficacy gaps seem to emerge within various underrepresented populations, previous research fails to provide an explanation of root causes from the perspective of the children and/or adolescents. In the following sections, I describe the design and other elements of this study aiming to identify those root causes. Research Design Qualitative research design was applied in this study. Qualitative research methodology used include: one on one interviews, grounded theory, and phenomenology as these would provide the opportunity to gather data from individual participants to better understand why there is a deficit in participation based on their individual experiences, perspectives and opinions. This research design provided depth and meaning to the experiences of the individual (Leavy, 2017). At the same time, it allowed me to generate robust and descriptive data that was used to build a better understanding of beliefs, attitudes, and feelings (Leavy, 2017) women and BIPOC individuals have with relation to STEM/CTE related areas in a 6th to 12th (secondary) grade setting. As I tried to piece together individual girls/young women’s and BIPOC experiences from their formative years, I believe that qualitative research allowed me to better understand the observed phenomenon by prioritizing individuals' subjective experiences (Leavy, 2017). Setting The sample for this study was taken from two schools from a public school district along the northern Wasatch front: one high school (HS) and one junior high school (JHS). Both locations are from a city school district (CSD) that has a much more diverse population than the surrounding county school districts (CTSD). Community differences are pronounced when comparing demographic, socioeconomic status, and overall levels of education between residents of both city district and county district populations (IES NCES 2020) (for more details, see Table 1). 15 Table 1 Site Statistics HS and JHS 20221 Site vs. State HS JHS State Math Proficiency 17% 12% 41% Language Arts Proficiency 33% 17% 44% Science Proficiency 22% 13% 44% Hispanic or Latinx2 50% 62% 18% Black Native American Pacific Islander 1% 1% 1% 1% n<10 1% 1% n<10 2% Two or More Races 3% 3% 3% Free Lunch Eligibility 45 % 99% 25 % Fifty percent of students at HS and ten percent of students at a high school in CTSD identify as Latinx (PSR, 2023). Community and student population demographics would indicate that sites within CSD provide a better participant pool for this study, which is where I recruited my sample from (for more details, see Table 2). Table 2 Hispanic/Latinx Proficiency Scores by Site3 Site HS JHS % Students Identifying as Hispanic or Latinx 50% 62% Hispanic or Latinx4 % Identifying as: Math Proficiency Language Arts Proficiency Science Proficiency Boys Girls 10% 8% 20% 10% 9% 8% 47% 52% 53% 48% Of the three total junior high schools located within CSD boundaries, JHS has the largest Latinx identifying student population (62%). JHS also has 99% of the total student population eligible for free lunch (PSR, 2023). The state average for Latinx identifying individuals in all public schools is only at 18%, while only 25% of the total state student populous is eligible for free lunch (PSR, 2023). JHS’s total student population has an average math proficiency score of 13%, reading proficiency at 17%, and science proficiency at 13%, which is the lowest 1 (PSR 2023; USBE 2022) Hispanic and/or Latinx are the racial/ethnic terms used in the demographic information provided by PSR or USBE. 3 (PSR 2023; USBE 2022) 4 Hispanic and/or Latinx are the racial/ethnic terms used in the demographic information provided by PSR or USBE. 2 16 proficiency for all three subjects of the junior high schools within CSD (USBE, 2022). Latinx identifying student proficiency in math, reading and science at JHS are 8%, 10% and 8%, respectively (USBE, 2022). HS, on the other hand, has a 50% Latinx-identifying student population with 45% of the total population eligible for free lunch (PSR, 2023). Math proficiency scores, based on state mandated testing, at HS are at approximately 17%, science proficiency at 22% and reading proficiency of 33% (USBE, 2022). Latinx identifying students at HS score a proficiency for language arts, math, and science at 20%, 10% and 9%, respectively (USBE, 2022). Of the 1036 students who attend HS, 47% identify as boys/young men and 53% as girls/young women (PSR, 2023). JHS has a student body of 703 students with 48% identifying as girls/young women and 52% as boys/young men (PSR, 2023). Sample For this study, I conducted nine (n=9) individual one-on-one interviews with nine students that are within the traditionally underrepresented STEM population (subjects that include: computer science, engineering, and physics), which according to literature, are women and BIPOC identifying. Based on teacher input and referrals, my intention was to select students that identify as girls/young women and/or BIPOC from the age range of 12 to17 years from junior high and high school populations. The rationale for this decision was the fact that schools in the middle to upper high school levels offer a variety of coursework related to STEM fields, whereas elementary schools vary in extracurricular offerings and would not have provided consistent enough comparable data for this project. In addition, literature shows that students in the above specified age range already have established stereotypes regarding STEM fields (e.g., Blazev et al., 2017; Cvencek et al., 2011; Grossman & Porche, 2014; Ing & Nylund-Gibson, 2013; Master et al., 2021), which may influence their choice of STEM-related subjects. These inculcated adolescent stereotypes provided an opportunity for discussion and data collection during the interviews. 17 Underrepresented students were selected based on the following criteria: ● 10th or 11th grade students currently enrolled in physics, computer science and/or engineering coursework. ● 8th or 9th grade students currently enrolled in physics, computer science and/or engineering coursework. ● Students who identify as girls/women or BIPOC were given priority for interviews. Students were provided with a demographic survey to compete prior to the start of the interview (see Appendix B). The survey was prepared and given via Google Forms. Data was recorded automatically into a spreadsheet exported from Google Forms. Table 3 provides some demographic information from the sample based on the given survey. Table 3 Demographic Information Participant Pseudonym Grade Gender5 Age Race and/or Ethnicity6 as marked by the participant2 Ashley 8th 13 Female Hispanic or Spanish, White Liara 8th 13 Girl Hispanic or Spanish Tali 8th 13 Female Hispanic or Spanish, White Jane 8th 14 Female Hispanic or Spanish, Latinx Jack 10th 15 Female White Aria 10th 15 Female White Samantha 11th 16 Female White Miranda 11th 16 Female Latinx, White Kelly 12th 17 Female White 5 6 Gender was presented as a write-in response rather than check box. Race and/or Ethnicity was presented as a checkbox with an additional option to write in a response. 18 Recruitment For this study, I used purposeful sampling (Patton, 2002) with the help of teachers of the subjects of math, science, and computer science or engineering from the above-mentioned schools. Purposeful sampling is a technique widely used in qualitative research for the identification and selection of information-rich cases for the most effective use of limited resources (Patton, 2002). Upon receiving approval to begin this project from CSD, l compiled a list of teachers from all schools that taught either: science, math or CTE coursework. I identified and contacted teachers (n=62) of these subjects based on current website data, although some position assignments were no longer accurate at the time of recruitment. I emailed each of the 62 teachers individually stating the purpose of my project and how they could assist me in my research. I also provided a short video via YouTube where I introduced myself and outlined the goals of the project in a teacher/student friendly manner. I received several responses from teachers indicating willingness to assist in this project. I further communicated my desires to select students that fit the traditionally underrepresented populations with the specific STEM fields of physics, computer science and engineering. This methodology is in accordance with snowball sampling as it relies heavily on teacher recommendation for student participants as they know their students or know of students whereas, I do not (Parker et al., 2019). I also expressed wanting to talk to any students that they felt could provide appropriate feedback during an interview with someone who is unknown to them (Parker et al., 2019). I also recommended that if there were no specific students in mind, that they display a pre-recorded introductory video and disseminate the information to applicable courses they taught (aka physics, chemistry, engineering, math, and computer science). This is in accordance with purposeful sampling methodology (Patton, 2002). The primary method of student recruitment was done through teacher recommendation (students from HS); however, the use of the video introduction was also used in classes where the teacher did not have any specific students in mind (students from JHS). I also made myself 19 available should the teacher request I come and speak to the class in person. This is in accordance with snowball and purposeful sampling, as it allowed me to find good participants by surveying people that have a vested interest in student achievement and are familiar with student interests and coursework (Patton, 2002), while the video provided an opportunity for all students in math, science, or computer science coursework to participate without teacher input should they choose. Teachers that responded via email that they had interested students in their classes were provided with informed consent forms (Appendix C) as well as an informational letter (Appendix D). The teacher provided the student with the consent/ascent form and informed me via text or email when the consent/ascent form was returned. All completed consent/ascent forms were left for pick up in the JHS or HS main office in an enclosed envelope. I picked up forms promptly from the school’s secretary after which I scheduled interview times with the teacher. Interviews were scheduled during the recruiting teacher’s class period with the student to minimize time the student will miss from classes. An hour was allotted for each interview, while times and dates were approved by the participating teachers. Students were compensated for their participation in the form of a $20 gift card to Target. Data Collection and Procedure The city school district (CSD) I collected student data from has their own protocol requirements for completing research within their district. Upon approval of the IRB from Weber State University, CSD was contacted for approval to conduct interviews within their junior high and high schools. Approval took several weeks to return. After I received approval from both WSU and CSD and received consent/ascent forms from each parent/student, I scheduled interviews at each school by contacting and working with that student’s participating teacher, as described above. 20 At JHS, I conducted interviews in an empty office inside of the school’s counseling center. Each of the four interviews took place in the same location. At HS, interviews took place either in the assistant principal’s office or the counselors’ office, depending on the availability of the spaces. Conducting interviews in small offices provided a better environment for the student to focus on the questions as well as to limit interruptions by staff or other students. At the beginning of the interview, I used Google Forms to collect demographic data. Students used a laptop to record their demographic information. Students were not required to sign in to complete the survey via Google Forms. Once the surveys were completed, I transferred the data to a spreadsheet for the assigning of anonymous student identifiers (see Table 3 above). The Voice Recorder app was used to record interviews and audio files were uploaded to Google Drive after which Otter.ai was used to create a baseline transcription. Transcriptions were verified and edited for accuracy of dictation and were then exported to a word document for further analysis. Students were interviewed on an individual basis one by one and asked a series of questions (see Appendix A) based on their personal experiences in their education. Each interview was audio recorded using a voice recorder app on my cell phone. Object based elicitation (OBE) (Barton, 2015; Harper, 2002) was utilized to gather data on student perceptions of topics, which included a forces free body diagram, the periodic table, a blockstructured programming language example, an example of a general-purpose programming language like Python, as well as the definition of a derivative from calculus (see Appendix E for examples). For qualitative research that involves interviewing subjects, OBE has been shown to be a powerful tool in eliciting feelings, memories, and pertinent information related to gathering opinions and attitudes for qualitative research interview questions (Barton, 2015; Harper, 2002). I printed each of the photographs on card stock and in color, providing one set per student (see 21 Appendix E). The five OBE cards were all placed in an identical starting order for each student. I numbered each of the OBE cards during each interview, after the student placed them in order with which they felt the content of the cards were easiest to the most difficult, one being the easiest or simplest and five being hardest most difficult. Data Analysis The analysis of collected data was done through open coding and axial coding. Open coding is recommended for primary identification of concepts and review of the data corpus (Saldaña, 2015). It “involves discovering patterns, themes, and categories in one’s data” (p.453) and is typically done in the early stages of data analysis to develop manageable classification or a coding scheme (Emerson et al., 1995). This stage of analysis involves identifying and naming important occurrences and their significance from the recorded transcripts. To secure transparency of the participants voices in the study, initial coding was employed in combination with In Vivo Coding wherever possible. Saldaña (2015) defines In Vivo Coding as coding that “draws from the participants’ own language for codes” (p.84), which was particularly useful in describing the culture and other perceptions that could be seen as integral to women and BIPOC persistence STEM fields. This phase resulted in a total of 196 codes, after which I proceeded to the second cycle of coding. The goal of the second cycle of coding was to understand the relationships between the codes identified in the first cycle and consequently, identify the most common themes (Saldaña, 2015). In the second cycle of coding, I employed axial coding, where I focused on organizing codes into related categories (Patton, 2002) which best answered the RQ. By moving from open to axial coding, dominant and relevant patterns, marked as categories, were identified, and organized to best answer the proposed research question (Patton, 2002). This coding cycle contained several iterative cycles of review and revision by myself and my advisor to make sure that coding is consistent with the proposed themes. As two researchers worked on reaching consensus over all the themes, as well as which codes go 22 within each theme, we did not calculate an additional calculation of intercoders’ reliability. At the end of the second cycle, 196 codes were organized into four categories, all aimed at answering the RQ (see Figure 1 for frequency of codes per theme/category). The themes that emerged from this analysis are: Family Matters, “She actually knew, like how to explain math”, “I don’t think I could code to save my life”, and “all we do is worksheet’s”, Table 4 outlines the codebook for each category that emerged. Each category has a definition and examples. The categories themselves are described in detail in the Results section and discussed in the final section, Discussion and Conclusions. Names used in the table are pseudonyms. Figure 1 Frequencies of Codes 23 Table 4 Codebook Theme “I don’t think I could code to save my life.” N=88 S=9 “She Actually, Like, Knew How to Explain Math” N= 64 S=9 “All we do is Worksheets.” N=35 S=8 Family Matters N=8 S=4 Definition Difficulty and/or perceived difficulty of subjects play a role in student opinions as attitudes regarding experiences and/or future opportunities. Aspects related to how a subject is labeled by the student arises from: prior subject experiences, selfefficacy/confidence regarding subjects, and having to work alone online vs. being in the classroom with access to a teacher. Teacher influence has a positive or negative effect on student attitudes, opinions, and retention. Student engagement directly relates to the concept of subject matter being perceived as boring or interesting. This category represents aspects of student experiences that relate to family support. Example Miranda stated regarding chemistry: “It was just a difficult subject for me to wrap my head around.” Jack, when asked about taking a hypothetical physics class, “I feel like I’d be a little bit intimidated.” “He (math teacher) would just give us lesson plans and like, just kind of throw it at us on a computer. And you just have to figure it out. Like it sucked. It really did suck trying to like figure everything out”. When asking for help, Jane stated “I did, but he (coding teacher) didn’t explain it well.” Ashley “…last year science, I didn’t really like it and it’s like, boring. I kind of find it useless.” “(My) cousin was joining it. And she asked me if I wanted to try and I was like, yeah, sure why not?” (regarding attending an afterschool coding club). “(My dad) he said that what I think architecture is, isn’t as much fun as I think it is”. (regarding participants career aspirations of becoming an architect). 24 For data pertaining to the OBE activity, during each interview, I labeled the cards with the numbers one through five, five being most difficult, one being the easiest in the students’ opinions. I imported OBE data based on each student’s response into Excel spreadsheet and created a chart (Figure 2) that shows the variation in response as well as patterns of perceived ease or difficulty based on student perception of the concept being displayed on each card. Figure 2 Student Ranking of OBE with Number of Responses Based on Ranking Results A sample of nine students completed the demographic survey for this study (see Table 3 for more demographic data on the participants). All the participants from JHS were in eighth grade, and they make up 44% of this study population. Participants from HS were in grades 10th-12th comprising 56% of the study population. Ethnically, 56% of the study group identified as Hispanic or Latinx and racially, 78% identified as White. All participants identified their 25 gender as girls/young women. The ages of participants ranged from 13 to17 years with the mean age being 14.7. As a reminder, my research question focuses on examining student feelings and opinions about STEM subjects, particularly math, physics, and computer sciences, such as programming (coding). After analyzing the data, four themes emerged as important root causes of STEM disparities among K-12 students from their own perspective. Themes are broken down in the following sections and they highlight common attitudes, responses, and reasoning that the participants presented during interviews. I don’t think I could code to save my life The most frequent theme that became apparent during the interviews was the idea that certain subjects are more difficult than others. This was stated regardless of if a participant had taken a so-called “difficult” class or if they simply had the impression or perception that it was a difficult class. All nine participants (code n=65) remarked about the difficulty they felt a certain course had, particularly regarding coding, physics, and math (calculus). Some courses were identified as difficult based on what others had stated to them while others became difficult due to the nature of content delivery: online format due to COVID or teacher preference. Generally, courses identified as difficult were predominantly in the sciences: chemistry and physics (n=9), while computer science was deemed overall difficult by both the JHS and HS school participants (n=6). Specifically, these participants had taken one or more of the mentioned “difficult” classes and regarded them as “having a hard time.” Other students stated they would feel intimidated to take certain classes within the STEM fields and one commented “I don’t think I could code to save my life” when looking at the line code for the object-based elicitation activity. Among the participants in this study, there was a prevailing idea that certain subjects are inherently more difficult than others. As a subtheme to this code, several participants (n=14) identified that physics and calculus were the most difficult subjects (from OBE), as well as 26 stating that they found it “looked” difficult. This also seemed to be the case with ranking the line coding in the OBE activity. Only one of the nine participants stated they had taken a physics class prior to the interview, but almost half stated that they considered physics to be nearly the most or the most difficult of the concepts covered in the OBE exercise. Three participants mentioned they struggled with online learning. In an online learning environment as well as in accelerated classes or classes where the subjects were thought to be “hard”, participants expressed frustration with “having to keep up” and having to “…figure it out” outside of class time and without teacher availability or adequate support. Students who had taken a class that they labeled as difficult attributed much of the perceived and/or experienced difficulty to the lack of teacher support. Other participants struggled with subjects that were new or were “[difficult to] wrap their head around.” For this phenomenon, coding came up among several (n=6) participants as “hard” or “tough” to feel successful in. One participant stated: “I think it’s just too hard for me to like, to understand. And it’s just hard for me to understand like, what do you do with this, like what the program, what it involves.” Though this unique situation came from one individual, it was repeated in similar phrasing by other participants about coding and/or hard sciences like chemistry or physics. Self-efficacy was a common sub-theme mentioned or discussed at various levels by all nine participants (n=22). The descriptions of being overwhelmed or not good at a particular subject dictated much of the feelings surrounding that subject. This mostly occurred when math was brought up by either me or the student. Being overwhelmed seemed to also hint that the amount of homework and feeling of needed “good grades” being factors in whether a subject was accessible to the student. Math, science (general) and coding were all mentioned as courses with which a participant felt overwhelmed, not particularly good at it or did not understand it enough to feel like they’d want to take another class in that subject. 27 Prior experience with certain courses came from a similar theme as teacher influence and self-efficacy, however, the difference being that this category involved no mention of teacher or how a student rated their ability/perception of ability in the discussed subject. Often comments delved into a participant’s initial experience with a subject or concept and later, they describe how their opinion changed for the positive or that they became disengaged toward the course. This theme involved discussion of prior coding experiences and several participants mentioned coding in elementary school to compare to their current or most recent experience with coding in junior high school. Comments about prior coding experiences range from “when I first started doing it, I didn’t really (like it)” to “I think it would be a good idea to take another one, I just don’t (have) a lot of interest in it”. Comparisons were also made between the type of coding program being utilized in prior classes being simpler (e.g., drag and drop boxes) versus what is being done at the middle school level. One student remarked “I took one in junior high. It was just awful. I don’t remember the website, but it was like the little blocks you drag.” Whether this was due to the type of coding being done or the impact of the teacher was not directly stated, only that their experience was not a positive one. A few participants commented that they may have enjoyed a class more had they had friends in the class or if the class had not been taught online (COVID or other at home learning days). She Actually, Like, Knew How to Explain Math According to the students in this study, teacher influence played no small part in student perspectives on subject matter. Both positive and/or negative influences of teachers were identified by all nine participants in the study. When it comes to math, science, and coding, a teacher’s positive or negative influence was reported to have the power to sway student opinions on difficulty, enjoyability and resilience in the subject. Of the 196 codes collected from nine participants, 61 codes (31%) directly referenced a teacher regarding math, science, or coding (computer science), making this theme the second most prominent in the data. 28 A recurring subtheme within this category seems to be a teacher’s explanation or lack of explanation regarding questions that arise during instruction. Several students pointed out that the acquisition of a particular topic was hindered by teachers’ lack of clarification when questioned. Several students alluded to the fact that they were not given proper explanation that would have allowed for them to firmly grasp a concept being presented, leading them to believe that the subject was difficult or unobtainable. The reverse also seems to be true, as several students responded that they felt when a teacher explained a subject well, they would begin to understand it, and it would begin to “make sense”. Within this same theme comes the influence of a teacher on students taking other subjects or classes in future semesters/grade levels. This had both a positive and negative influence on the students’ perspective. For example, when asked if a participant would take a coding class based on a particular teacher’s suggestion, one student responded, “I trust most of my teachers and their general decisions that I feel like, I feel like I could handle it.” Other comments regarding future classes were less positive. One participant stated that “I don’t know (regarding taking physics in the future). Just like, like teachers complain about having to take physics in high school, hating (sic) and like, makes me a little nervous.” These are two starkly different perspectives and provide a look at how teacher influence in a subject can be considered a root cause for whether a student would or wouldn’t take a coding or physics class as a future elective. Other common themes within this code identify that teachers that are seen as “effective” in their methodology and approach to teaching are seen as having more enjoyable classes, even in subjects that may not have been a favorite of the participant. Participants stated that teachers who “made sure like, we got it down” and “just had a really effective way of teaching” were favorite educators/classes. The opposite also seems true regarding student attitudes and perspectives of coursework. One participant identified her least liked class was a math class where the teacher “didn’t go into the details and help the students like, make sense of it.” 29 All we do is worksheets Seven of the nine students (code n=30) commented on the importance of engagement, either for the negative, as in the subject is considered “boring” or for the positive, the subject is a “favorite” or considered interesting. Science was mentioned the most when it comes to subjects considered “boring” or labeled as hated. Comments ranging from “I hate taking notes” to “I find it useless” were all directed at middle to high school science courses. This lack of engagement directly influenced whether a student found it valuable or not. A common complaint seemed to stem from the type of work being conducted in the science courses. Comments ranged from taking notes to having to complete worksheets and the lack of hands-on projects. Several students commented that projects would have contributed to a more enjoyable, engaging class. Science projects were recalled favorably and even remarked as being a favorite experience for some participants. Family Matters The impact of family support discussed by four of the nine participants (code n=8), focused on grades, the types of course work being taken (AP vs. regular classes), and career aspirations. Participants commented that parents or other family members would assist with homework and encourage attendance in classes and school. Pressure from parents to perform well or get good grades seemed to also place pressure on one participant to perform well in school. In doing so, this participant sought to drop from an AP science class to a regular class where they felt they would have more success grade wise. Another student mentioned that their career goals were minimized and questioned by a parent, leading them to second guess their aspirations in that field. Discussion My research question focuses on examining student feelings and opinions about STEM subjects, particularly math, physics, and computer sciences, such as programming (coding). Based on the interviews and subsequent axial coding, I was able to identify four major themes 30 that help identify possible root causes of deficit participation in the focus STEM subjects: math, physics, and computer sciences: Prior Experiences, Perception of Difficulty, SelfEfficacy/Confidence and Willingness to Engage. These four themes or root causes appear to correlate with each other, one leading to the formation of the others. Overall, prior experiences based on teacher influence and initial experience correspond with student efficacy and confidence in subjects as well as their perception of how the subject may or may not be considered difficult. These inner connected root causes then seem to influence a student’s willingness to engage further with that particular subject in the future and seem to shape the students’ overall attitude regarding that subject. Figure 3 Correlation Between Root Causes Prior Experience Research suggests that teachers have a lasting impact on student attitudes and perceptions, both positive and negative (e.g., Hattie, 2011; Master et al., 2017; Skinner & Belmont, 1993; Sullivan & Bers, 2018; Walan, 2019). As prior studies show, the facilitator and/or 31 educator involved in delivering interventions can have a positive effect on student attitudes, even going as far as to change preconceived notions about STEM subjects like computer programming (Master et al., 2017; Sullivan & Bers, 2018). For the participants in this study, the impact of teachers, both prior and current, was tantamount to developing their feelings of selfefficacy, confidence, and perceived adequacy in any subject. A teacher’s influence on students’ confidence in a subject has been well documented in prior studies (e.g., Niepel et al., 2019; Pantic, 2020). Data from this study shows that teachers who taught subjects “well” were regarded favorably and seemed to have a positive impact on students pursuing classes in that subject in future grade levels or semesters. When it came to subjects like math, science and coding, the opinions and attitudes surrounding these subjects were generally more favorable toward that subject if students had good rapport with their teachers or considered them to be “good” teachers. This is supported by prior research that had similar findings (Olszewski-Kubilius et al., 2017; Skinner & Belmont, 1993). In contrast, those participants stated that had a negative experience in coding, math or science class and had much less interest in taking more than was required of these subjects. This is aligned with higher education literature on women’s retention which found that women, who are critically underrepresented in computer science, are much more likely to report negative experiences with or opinion about the computer science faculty (Barker et al., 2009; Denner et al., 2014), In his book on visible learning strategies, John Hattie (2011) describes some potential ways teachers and classroom environments can have both equally positive and negative effects on student achievement. Reciprocal teaching and peer collaborative have a highly positive effect on student learning (Hattie, 2011), while limiting such interactions can be somewhat detrimental to a student’s ability to make sense of “difficult” material. Other study participants explained that when a teacher just tells you how to do it rather than showing them where to start or explaining things clearly, made the subject more difficult to understand and eventually they just did not enjoy the class or course. These negative experiences apply to both math and 32 science as well as coding and led to a poor perception of these subjects whereas teachers who spent time working toward comprehension with students had a much more positive assessment of their teaching ability by participants which is in line with Hattie’s research (2011) on how student-teacher relationships impact learning, attitude, and achievement. My study adds nuance to those insights. For example, teachers that were regarded favorably were teachers that made them “feel welcome,” had a positive attitude about school and took the time to show them how to do math and science rather than just “telling” them how it’s done. I do believe this aspect of the interviews addresses a root cause directly in that teaching techniques and teaching styles have a lasting impact on student achievement and feelings of self-efficacy as teacher/mentor rapport has been shown in prior research to have generally a high effective size on student attitudes and achievement (Hattie, 2011; Niepel et al., 2019; Pantic, 2020). Engagement can be crucial to holding a student’s interest and attention in any subject (Daniels, 2011; Hattie, 2011). When it comes to STEM subjects, particularly the ones in the focus of this study, engagement becomes imperative as stereotypes are already hardwired into student’s’ perceptions of subject matter (Master et al., 2017; Olszewski-Kubilius et al.,2017). Based on student interviews, engagements seem to be part of the root causes for why students have a generally negative opinion regarding math, science, and programming courses. Many participants (seven out of the nine) complained that while also being difficult, math and science courses tend to also be “boring”, and the coursework is often associated with just completing worksheets and taking notes. Engagement looks different depending on grade level. As referenced in the prior section, the teacher is one of the most critical components influencing student engagement (Hattie, 2011; Skinner & Belmont, 1993). Generally, according to the participant interviews, science was the most “boring” of the subjects while coding was considered something that was pointless or not much was to be taken from the class other than needing it for credit purposes. According to 33 the participants from this study, most of these classes are work that involves mostly teacher talk and provide little hands-on opportunities to apply what is being taught in a real world or working scenario. This is further made apparent by participants stating that they would much rather have activities and experiences in their sciences courses, and that hands on experiences would add enjoyment and value to what they perceive as doing/learning science. Perhaps this highlights a major flaw in how science, math, and coding are presented to students in a middle to high school environment. If science, math, and coding classes are seen as things that are separate from real world scientific, mathematical, and programming endeavors, then I can understand why students would have two very different ideas of what these subjects look like. A prior study by Kontra et. al. (2015) suggests that hands on learning, particularly in science, saw improved performance academically as well as higher engagement. This study also concluded that regions of the brain responsible for sensorimotor were activated which could enhance student acquisition of difficult concepts (Kontra et.al., 2015). With the implementation of the newest state standard, there is a direct emphasis on hands on application for science learning. Whether this is being implemented in the classroom, however, is not clearly evidenced as student data suggests it is not the norm. Perception of Difficulty and Self Efficacy The most prominent theme identified through the interviews of the nine students was the perception of or general feeling of difficulty associated with certain types of math, sciences, and computer science (coding). All nine participants remarked that these courses are often causes of anxiety related to content homework and testing due to their varying degrees of difficulty either experienced by the student or perceived by the student. This is tightly related to Bandura’s self-efficacy theory (Bandura, 1977). A lot of research agrees that self-efficacy in a subject plays a part in the perception of a subject’s innate difficulty (e.g., Ball et al., 2019; Clark et al., 2018; Hattie, 2011; Stoeger et al., 2013; Walan, 2019). Students who felt confident in a 34 subject stated that they liked the subject, felt it was easy for them and that they enjoyed classes in that subject. The opposite to this, is when students felt they struggled or just “didn’t get it”. These selfassessments and opinions of self-adequacy play an integral part in a student’s desire to continue in a subject (Ball et al., 2019; Clark et al., 2018; Stoeger et al., 2013; Walan, 2019). The feelings of being overwhelmed in math, science and programming were reiterated by multiple study participants. Being “good” at a subject seemed to signal to students that it meant that the subject was “easy” or at least easier for them, while subjects that they were “bad” at or struggled with were inherently “hard” (Charlesworth & Banaji, 2019; Kurtz-Costes et al., 2014; Niepel et al., 2019). Much of what was deemed “hard” by students dealt with not being able to understand something right away. The initial introduction of a subject appeared to have provided a baseline for how these students judged the subject’s ease of acquisition. This is aligned with Bandura’s (1977) concept of mastery performance. The initial experience with a subject directly influenced students’ rationale for taking a repeat class in that subject and whether they felt the subject was easy or difficult, which is similar to other studies I found (e.g., Blazev et al., 2017; Blue et al., 2018; Charlesworth & Banaji, 2019; Hattie, 2011; Khan & Rodrigues, 2017). Certain subjects seemed to have an innate “intimidation” factor to them as indicated by the participants. Physics for example, was mentioned by several students as being “difficult” despite not having taken coursework in a physics class before. Coding, on the other hand, was stated to be a difficult class for many of the participants, while block coding was ranked as the easiest by almost all students. Such disconnect between coding being ranked both easiest and most difficult seemed to be based on the visual representation of coding. For most, it was not coding that seemed to intimidate, it was the representation that caused students to feel that it (line/python coding) would be too difficult for them. 35 This link between visual representation and interest and/or difficulty seems to be rooted in the prior/initial experiences and the attitudes cultivated because of those initial experiences (Bandura 1977; Hattie, 2011). These attitudes are reflective of other studies done where selfefficacy, interest and general opinions of subjects are directly influenced by having prior experience with positive role models (teachers, mentors, peers etc.) (e.g., Bandura, 1977; Clarke-Midura et.al., 2020; Clarke-Midura et.al., 2018; Hattie, 2011). It is, therefore, important to take those considerations into account when designing students’ first experiences with these subjects. Participants also attributed difficult to “having to figure it out yourself.” Some courses, according to the participants in this study, particularly during the pandemic school closures, became more difficult for students to receive help with questions and adequate access to their instructors. However, this is not isolated to pandemic shut down situations. Online learning and the feeling of working alone also seems to occur in classes where it seems teachers are reluctant to spend time making sure students are comfortable with their understanding of a subject (e.g., De Paepe et. al., 2018; Dhilla, 2017; Lachlan et. al., 2020; Leech et. al., 2022;). This link between prior experiences and their impact on perceptions of difficulty of a subjects seems to play a definitive role in whether a student will seek to take further classes in that subject (Bandura, 1977; Hattie, 2011). In the case of computer science (programming) many students spoke negatively of taking a repeat class, stating that they had a negative prior experience and/or thought the material was too difficult, due in no small part to not having access to teacher support and/or not being fully assisted by an instructor in a way that made the student feel their questions and concerned were answered properly. Teacher influences also play a key role in inculcating students’ attitudes and opinions of STEM subjects (e.g., Ball et al., 2019; Clark et al., 2018; Hattie, 2011; Stoeger et al., 2013). I will discuss these implications in the next section. 36 Family support is generally regarded as a positive factor in the participants’ educational endeavors (Henderson & Mapp, 2002; Lin, 2016). A small sample of the participants (n=8) mentioned having supportive family systems that assisted in their selection of courses, encouraged good attendance, and emphasized the importance of good grades as well as going to college. Obtaining good grades seemed to be the most important and most frequently mentioned aspect of familial support, adding that receiving poor grades would make parents upset. Though this type of support can be seen as a positive factor for students, it seems to add to the feelings of anxiety and being overwhelmed those participants mentioned when discussing the perceived/apparent difficulty of some classes. For example, several students mentioned the need to get good grades. When asked to clarify what “good grades” looked like, participants advised that only “A’s” were considered good grades to their parents/families. If obtaining only ‘A’ grades in all classes is expected of students, then it seems that taking classes that are thought of as “difficult” to comprehend would not be desirable and therefore, would not be taken if not required. Familial expectation of high performance may indeed be a root cause in the sense of the pressure for students to achieve certain academic standing becomes more important than the types of classes they take. This relates back to the main codes or themes identified in this study, that is the difficulty of a subject may be directly related to the teacher’s ability to engage and inspire students which in turn, effects whether students will seek to retake or continue in those subjects. The Willingness to Engage The willingness to engage with coursework beyond what is required, particularly with math, physics, and computer science, seems to be directly related to the students’ initial experience with the teacher, the subject matter difficulty (actual or perceived) and the engagement they felt with their prior experience in that subject as well as a student’s own 37 perception of self-efficacy. These “root causes” have a direct effect on a student’s self-concept and thus affect whether they would choose to engage further with that subject. Students who reported having positive, engaging experienced were more likely to take a future class or extension of that class whereas students who reported having a negative experience with the teacher, the material or engagement with the subject were less likely to take a future class, even if the possibility was presented as hypothetical. These identified possible root causes and their relation to retention/persistence within certain STEM subjects seem to provide a better understanding at where the deficit may be stemming from and can further shape our endeavors to close the gap in future work. Conclusions In seeking to identify the root cause(s) of deficit participation in STEM by women and BIPOC populations, this study focused on looking at student perceptions and attitudes regarding the STEM subjects of math, science, and programming (coding). Participants from both the middle and high school levels each expressed criticism over the difficulty of such subjects, the lack of engagement in such subjects as well as the perception that the subject was too hard or that they were not comfortable with its concepts. A major critique of teacher efficacy came into question regarding science and coding coursework at both the junior high and high school levels. Teacher influence and a teacher’s ability to make a subject more engaging seems to be important factors in student persistence in math, science, and programming. Student’s own perception of self-efficacy can be greatly influenced by a teacher, and thus, teachers seem to bear the responsibility of making these subjects comprehensible as well as providing hands on and real-world application so that students involved in these courses can see the benefit and usefulness of such coursework. The feelings of difficulty and self-efficacy, from the student perspective, seem to point toward teacher effectiveness and their ability to engage students, even when dealing with 38 content that traditional stereotypes would label as “difficulty”. While family is important to overall student success, it seems, based on participant responses, that it also adds pressure to perform and thus may also influence the type of coursework a student would choose to take. These factors/root causes all seem to allude to deficit participation being a multifaceted issue that begins with a student’s initial experience in a subject (Bandura, 1977; Hattie, 2011). Student experiences are shaped by teacher efficacy and engagement within that subject and perceptions of difficulty within that subject are a direct result of those prior experiences (e.g., Ball et al., 2019; Clark et al., 2018; Hattie, 2011; Niepel et al., 2019; Pantic, 2020; Stoeger et al., 2013). Thus, a student is more or less likely to engage further with that subject based on these root causes. Limitations This study has its limitations. The first is the sample size is small. Having only nine participants does not provide nearly enough data to conclusively identify root causes. The sample size did not include all marginalized or underrepresented groups in the fields of mathematics, hard sciences (physics) and computer sciences. The second limitation the sample contains only the perspectives of girls/young women from White and Latinx/Hispanic backgrounds as they were the only subjects that applied to participate in this study, leaving other BIPOC populations and their perceptions unrepresented. Third, all participants from JHS were all from the same math class and students from HS were all from the same architecture class. This limits the sample population to just two small pools of applicants and may have led to certain shared experiences and attitudes being reiterated versus having a more diverse sample from multiple courses and teachers from across the district/school sites. Implications and Future Work As stated in the discussion, this study sets the stage for future research into the root causes and how to effectively address issues presented in this study. Going forward, steps can be taken to address engagement in subjects identified by participants as “boring” or “useless”. 39 While the sample size was small, this study does set up future work with a larger and more diverse sample size that can shed better light on what best defines a “good” teacher as well as an “engaging” environment in math, science, and computer science. Gaining a more diverse perspective is also paramount for future work regarding some of the issues brought about by this study. Better defining the perspective of other BIPOC as well as gender inclusivity can also add important data to this study and provide more implications for how each root cause may better relate to one another. 40 References Accenture. (2020). The hidden value of culture makers. https://www.accenture.com/content/dam/accenture/final/a-com-migration/thoughtleadership-assets/accenture-getting-to-equal-2020-research-report.pdf#zoom=40 Ahmed, M. (2019). The good, the bad, and the ugly: implications of ethnic diversity in the workplace. Aisthesis, 10(1), 10-17. Barker, L. J., McDowell, C., & Kalahar, K. (2009). Exploring factors that influence computer science introductory course students to persist in the major. Proceedings of the 40th ACM Technical Symposium on Computer Science Education, 153–157. https://doi.org/10.1145/1508865.1508923 Barton, A. C., Tan, E., & Greenberg, D. (2017). The makerspace movement: Sites of possibilities for equitable opportunities to engage underrepresented youth in STEM. Teachers College Record (1970), 119(6), 1–44. https://doi.org/10.1177/016146811711900608 Barton, K.C. (2015). Elicitation techniques: Getting people to talk about ideas they don’t usually talk about. Theory & Research in Social Education, 43(2), 179-205. https://doi.org/10.1080/00933104.2015.1034392 Ball, C., Huang, K., Rikard, R.V. & Cotten, S.R. (2019). The emotional costs of computers: An expectancy-value theory analysis of predominantly low socioeconomic status minority students’ STEM attitudes. Information, Communication & Society, 22(1), 105-128. https://doi.org/10.1080/1369118X.2017.1355403 Bandura, A. (1977). Social Learning theory. Prentice-Hall. Boston Consulting Group (BCG) (2017) The mix that matters: Innovation through diversity. http://media-publications.bcg.com/22feb2017-mix-that-matters.pdf Briggs, C. (2017). The Policy of STEM Diversity: Diversifying STEM Programs in Higher Education. Journal of STEM Education, 17(4), 5-7. 41 References Blazev, M., Karabegovic, M., Burusic, J. & Selimbegovic, L. (2017). Predicting gender-STEM stereotyped beliefs among boys and girls from prior school achievement and interest in STEM school subjects. Social Psychology of Education, 20, 831-847. https://doi.org/10.1007/s11218-017-9397-7 Blue, J., Traxler, A. L., & Cid, X. C. (2018). Gender matters. Physics Today, 71(3), 40–46. https://doi.org/10.1063/pt.3.3870 Charlesworth, T.E.S., & Banaji, M. (2019). Gender in science, technology, engineering, and mathematics: Issues, causes, solutions. Journal of Neuroscience, 39(37), 7228-7243. https://doi.org/10.1523/JNEUROSCI.0475-18.2019 Cimpian, J.R., Taek, H.K. & McDermott, Z.T. (2020). Understanding persistent gender gaps in STEM. Science, 366(6497), 1317-1319. https://doi.org/10.1126/science.aba7377 Clarke-Midura, J.,Poole, F.,Pantic, K., Hamilton, M., Sun,C., & Allan. V. (2018). How near peer mentoring affects middle school mentees. Proceedings of the 49th ACM Technical Symposium on Computer Science Education, 664–669. https://doi.org/10.1145/3159450.3159525 Clarke-Midura, J., Sun, C., & Pantic, K. (2020). Making apps: An approach to recruiting youth to computer science, ACM Transactions on Computing Education, 20(4), 1–23. https://doi.org/10.1145/3425710 Cvencek, D., Meltzoff, A. N., & Greenwald, A. G. (2011). Math-gender stereotypes in elementary school children: Gender stereotypes. Child Development, 82(3), 766–779. Dhilla, S.J. (2017) The role of online faculty in supporting successful online learning enterprises: A literature review. Higher Education Politics and Economics, 3(1), 1-31. 42 Daniels, E. (2011). Creating motivating learning environments: Teachers matter: Teachers can influence students’ motivation to achieve in school. Middle School Journal, 43(2), 32–37. http://www.jstor.org/stable/23024496 Denner, J., Werner, L., O’Connor, L., & Glassman, J. (2014). Community college men and women: A test of three widely held beliefs about who pursues computer science. Community College Review, 42, 342–362. https://doi.org/10.1177/ 0091552114535624 De Paepe L., Zhu, C., & DePryck, K. (2018) Drop-out, retention, satisfaction, and attainment of online learners of Dutch in adult education. International Journal on E-Learning, 17(3), 303-323. Douglas, A.K., & Strobel, J. (2014). Hopes and goals survey for use in STEM elementary education. International Journal of Technology and Design Education, 25, 245-259. https://doi.org/10.1007/s10798-014-9277-9 Emerson, R. M., Fretz, R. I., & Shaw, L. L. (1995). Writing ethnographic fieldnotes. Chicago, IL: University of Chicago. http://www.press.uchicago.edu/ucp/books/book/chicago/W/bo12182616.html Estrada, M., Burnett, M., Campbell, A.G., Campbell, P.B., Denetclaw, W.F., Gutierrez, C.G., Hurtado, S., John, G.H., Matsui, J., McGee, R., Okpodu, C.M., Robinson, T.J., Summers, M.F., Werner-Washburne, M. & Zavala, M. (2016). Improving underrepresented minority student persistence in STEM. CBE-Life Sciences Education, 15(5), 1-10. http://doi.org/10.1187/cbe.16-01-0038 Fields, D. A., Pantic, K., & Kafai, Y. B. (2015). “I have a tutorial for this”: The language of online peer support in the scratch programming community. Proceedings of the 14th International Conference on Interaction Design and Children. Foma, E. (2014) Impact of workplace diversity. Review of Integrative Business and Economics Research, 3(1), 382-390. 43 Green, K.A., López, M., Wysocki, A., & Kepner, K. (2012). Diversity in the workplace: benefits, challenges, and the required managerial tools. University of Florida IFAS Extension, HR022. Grossman, J. M., & Porche, M. V. (2014). Perceived gender and racial/ethnic barriers to STEM success. Urban Education, 49(6), 698–727. https://doi.org/10.1177/0042085913481364 Harper, D. (2002) Talking about pictures: A case for photo elicitation, Visual Studies, 17(1), 1326. https://doi.org/10.1080/14725860220137345 Hattie, J. (2011). Visible learning for teachers: Maximizing Impact on learning. Routledge. Henderson, A. T. & Mapp, K.L. (2002) A new wave of evidence: The impact of school, family, and community connections on student achievement. National Center for Family and Community Connections with Schools, SEDL/AIR. https://sedl.org/connections/resources/evidence.pdf Hite, R., & Taylor, D. (2021). Fostering interest in and motivation for STEM: An illustrative case study of middle grade students’ experiences in out of school (OST) STEM activities. Journal of Interdisciplinary Teacher Leadership, 5(1), 1-23. https://doi.org/10.46767/kfp.2016-0035. Institute of Education Sciences: National Center for Education Statistics (IES NCES) (2020) District Demographic Dashboard 2013-2020. https://nces.ed.gov/programs/edge/acsdashboard Ing, M., & Nylund-Gibson, K. (2013). Linking early science and mathematics attitudes to longterm science, technology, engineering, and mathematics career attainment: Latent class analysis with proximal and distal outcomes. Educational Research and Evaluation,19(6), 510–524. Khan, Z.R., & Rodrigues, G. (2017). STEM for girls from low-income families: Making dreams come true. The Journal of Developing Areas, 51(2), 435-448. 44 Kontra, C., Lyons, D.J., Fischer, S.M. & Beilock, S.L (2015). Physical experience enhances science learning, Psychological Science, 26(6), 737-749. Krishnan, S.G. (2020) Gender diversity in the workplace and its effects on employees’ performance, Journal of the Social Sciences, 48(3), 1888-1898. Kurtz-Costes, B., Copping, K. E., Rowley, S. J., & Kinlaw, C. R. (2014). Gender and age differences in awareness and endorsement of gender stereotypes about academic abilities. European Journal of Psychology of Education, 29(4), 603–618. Kurtz-Costes, B., Rowley, S. J., Harris-Britt, A., & Woods, T. A. (2008). Gender stereotypes about mathematics and science and self-perceptions of ability in late childhood and early adolescence. Merrill-Palmer Quarterly (Wayne State University. Press), 54(3), 386–409. Lachlan, L., Kimmel, L., Mizrav, E. & Holdheide, L. (2020). Advancing quality teaching for all schools: Examining the impact of COVID-19 on the teaching workforce. Center on Great Teachers & Leaders at the American Institutes for Research. Leavy, P. (2017). Research design: Quantitative, qualitative, mixed methods, arts-based, and community-based participatory research approaches. Guilford Publications. Leech, N., Gullet, S., Cummings, M. H., & Haug, C. (2022). The challenges of remote K-12 education during the COVID-19 pandemic: Differences by grade level. Online Learning, 26(1), 245-267. Leu, K. B., & Arbeit, C. A. (2020). Differences in High School CTE Coursetaking by Gender and Race/Ethnicity. Career & Technical Education Research, 45(1), 33–61. https://doiorg.hal.weber.edu/10.5328/cter45.1.33 Lin, G.-Y. (2016). Self-efficacy beliefs and their sources in undergraduate computing disciplines: An examination of gender and persistence. Journal of Educational Computing Research, 53(4), 540–561. https://doi.org/10.1177/0735633115608440 45 Master, A., Cheryan, S., Moscatelli, A. & Meltzoff, A. (2017). Programming experience promotes higher STEM motivation among first-grade girls. Journal of Experimental Child Psychology, 160, 92-106. http://dx.doi.org/10.1016/j.jecp.2017.03.013 Master, A., Meltzoff, A. N., & Cheryan, S. (2021). Gender stereotypes about interests start early and cause gender disparities in computer science and engineering. Proceedings of the National Academy of Sciences of the United States of America, 118(48), e2100030118. Miller, D. I., Nolla, K. M., Eagly, A. H., & Uttal, D. H. (2018). The development of children’s gender-science stereotypes: A meta-analysis of 5 decades of U.S. draw-A-acientist studies. Child Development, 89(6), 1943–1955. Niepel, C., Stadler, M., & Greiff, S. (2019). Seeing is believing: Gender diversity in STEM is related to mathematics self-concept. Journal of Educational Psychology, 111(6), 1119– 1130. Okrent, A., & Burke, A. (2021, August). The STEM Labor Force of Today: Scientists, Engineers and Skilled Technical Workers. Science & Engineering Indicators. National Science Foundation [NSF]. https://ncses.nsf.gov/pubs/nsb20212. Olszewski-Kubilius, P., Steenbergen-Hu, S., Thomson, D. & Rosen, R. (2017). Minority achievement gaps in STEM: Findings of a longitudinal study of project excite. Gifted Child Quarterly, 61(1), 20-39. https://doi.org/10.1177/0016986216673449 Ozgen, C., Nijkamp, P., & Poot, J. (2017). The elusive effects of workplace diversity on innovation: Workplace diversity and firm innovation. Papers in Regional Science: The Journal of the Regional Science Association International, 96, S29–S49. Pantic, K. (2020). Retention of Women in Computer Science: Why Women Persist in Their Computer Science Majors (Doctoral dissertation, Utah State University). Pantic, K., Clarke-Midura, J., Poole, F., Roller, J., & Allan, V. (2018). Drawing a computer scientist: stereotypical representations or lack of awareness? Computer Science Education, 28(3), 232-254. 46 Parker, C, Scott, S & Geddes, A (2019) Snowball Sampling. SAGE Research Methods Foundations. https://dx.doi.org/10.4135/9781526421036831710 Patton, M. Q. (2020). Qualitative evaluation and research methods (3rd ed.). SAGE Publications. Public School Review (PSR) (2023) School and District Profiles. https://www.publicschoolreview.com/ Reinking, A., & Martin, B. (2018). The gender gap in STEM fields: Theories, movements, and ideas to engage girls in STEM. Journal of New Approaches in Educational Research, 7(2), 148-152. https://doi.org/10.7821/naer.2018.7.271 Saldaña, J. (2015). The coding manual for qualitative researchers. SAGE. Skinner, E.A. & Belmont, M.J. (1993) Motivation in the classroom: reciprocal effects of teacher behavior and student engagement across the school year. Journal of Educational Psychology, 85(4), 571-581. Sullivan, A., & Bers, M.U. (2018). Investigating the use of robotics to increase girls’ interest in engineering during early elementary school. International Journal of Technology and Design Education, 29, 1033-1051. https://doi.org/10.1007/s10798-018-9483-y Stoeger, H., Duan, X., Schirner, S., Greindl, T. & Ziegler, A. (2013). The effectiveness of a oneyear online mentoring program for girls in STEM. Computers and Education, 69: 408418. http://dx.doi.org/10.1016/j.compedu.2013.07.032 Utah State Board of Education (USBE) (2022) Data gateway. https://datagateway.schools.utah.gov/ Walan, S. (2019). The dream performance-a case study of young girls’ development of interest in STEM and 21st-century skills, when activities in a makerspace were combined with drama. Research in Science & Technological Education. https://doi.org/10.1080/02635143.2019.1647157 47 Yücel, Y., & Rızvanoğlu, K. (2019). Battling gender stereotypes: A user study of a code-learning game, “Code Combat,” with middle school children. Computers in Human Behavior, 99, 352–365. https://doi.org/10.1016/j.chb.2019.05.029 Zweben, S.H. & Bizot, E. B. (2016) Representation of women in postsecondary computing: Disciplinary, institutional, and individual Characteristics. Computing in Science & Engineering. 18(2),40-56. doi: 10.1109/MCSE.2016.21. 48 Appendix A: Interview Script and Protocol Interview Script: Hi! My name is Laura, and I am a student at Weber State University. I’m here to ask you some questions about math, science, and computer classes you may have taken since you started school and a few questions about school in general. There are not right or wrong answers, because I’m just looking for your feelings, opinions, and experiences. Answer as honestly as you can, please. Your answers will not be shared with anyone. I have some Legos here if you’d like to use them while we talk. Do you have any questions for me before we begin? Do you mind if I start recording? This recording will only be heard by me, and it is made to help me remember what we talked about. (Start recording) Interview Protocol: 1. Tell me about the times when you feel the happiest about being in your favorite classes? Why are these your favorite? 2. If you could take any class in the world, which class would you love to take? Why? 3. Which class would you never take? Why? 4. What classes are you taking now? Have you ever taken a coding, computer science or engineering class? 5. Tell me about your least favorite memory from a math, science, coding, or engineering class you took. 6. What is your favorite or least favorite things about your math classes? 7. (Object-based elicitation): I am going to show you a few photos. I’d like you to rank/order them from easiest to hardest. There isn’t a right or wrong way to organize them, so do the order based on how you feel. Can you tell me why you placed these the way you did? 49 8. If your teacher suggested that you take a computer programming class, how would you feel about that? Why? 9. Tell me about a time when you learned how to code/engineer something? How did it go? Did you have any experiences in elementary school with coding? 10. If your teacher suggested that you take physics as a science class, how would you feel about that? Why? 11. What kind of support for your education do you get? Do you feel like it helps you along the way? 12. Were you ever enrolled in a computer science, science, or math class that you really didn’t like? Why did you not like the class/decide not to take more classes in that subject area? 13. Is there anything else you would like me to know? About you, your classes, your experiences in school? 50 Appendix B: Demographic Survey Demographic Survey: This survey will be executed via google forms and available on a computer for the student to complete prior to the start of the interview: First and Last Name: What grade are you in: o 8th o 9th o 10th o 11th o 12th Your Age: What Gender best Describes You: (please write in) What race(s) and/or ethnicities do you identify with (you can select more than one): o Asian o Native Hawaiian or Pacific Islander o Black or African American o Native American or Alaskan Native o Middle Eastern or North African o Hispanic or Spanish o LatinX o White o Multiracial or Biracial o A race and/or ethnicity not listed (please write in) 51 Appendix C: IRB approval 52 53 Appendix D: Letter to parents Dear Parents, I am currently working on a Master of Education Degree in Curriculum and Instruction at Weber State University. For my final project, I have chosen what I consider to be a very relevant topic affecting today’s classroom. I am researching student perceptions and attitudes regarding the STEM fields of physics, math, computer science and engineering. To finish my Master’s degree, I am conducting a small research project that involves students within grades 8th-11th the Jr. High’s and High Schools in Ogden City School district. This study is very important and relevant to me and my experience as a former teacher as I was an elementary school science teacher and a STEM specialist for Ogden City Schools for thirteen years. I need, and would so appreciate, your signed permission to allow your student to participate in the study. The only information I will be using from your student will be demographic information as well as the interview itself. This information will be kept completely confidential. Your son/daughter’s identity will not be revealed to anyone not directly involved in conducting the research, nor will his/her identity be revealed in any publication, document, or computer database. I believe as teachers, we should always be improving our craft. The information gathered from this project will be useful in creating the most optimal learning environment for academic achievement and overall classroom experience for all students. Participation is voluntary and refusal to participate will involve no penalty or loss of benefits to which your student is otherwise entitled. To include input from your son/daughter in this study, we must obtain your written consent. If you are willing to have your child participate in this research project, please sign a copy of the enclosed Parent/Guardian Consent Form from Weber State University. Please sign and return if you consent to having your student participate in the study. Thank you and all the best, Laura Wilcock MEd Candidate and Student Weber State University 54 Estimados padres, Actualmente estoy trabajando en una Maestría en Educación en Currículo e Instrucción en Weber State University. Para mi proyecto final, he elegido lo que considero un tema muy relevante que afecta el aula de hoy. Estoy investigando las percepciones y actitudes de los estudiantes con respecto a los campos STEM de física, matemáticas, informática e ingeniería. Para terminar mi maestría, estoy realizando un pequeño proyecto de investigación que involucra a estudiantes de los grados 8 a 11 de las escuelas secundarias y preparatorias del distrito escolar de la ciudad de Ogden. Este estudio es muy importante y relevante para mí y mi experiencia como exmaestro, ya que fui maestro de ciencias de la escuela primaria y especialista en STEM para las escuelas de la ciudad de Ogden durante trece años. Necesito, y agradecería mucho, su permiso firmado para permitir que su estudiante participe en el estudio. La única información que usaré de su estudiante será la información demográfica, así como la entrevista misma. Esta información se mantendrá completamente confidencial. La identidad de su hijo/a no se revelará a ninguna persona que no esté directamente involucrada en la realización de la investigación, ni se revelará su identidad en ninguna publicación, documento o base de datos informática. Creo que como maestros, siempre debemos mejorar nuestro oficio. La información recopilada de este proyecto será útil para crear el entorno de aprendizaje óptimo para el rendimiento académico y la experiencia general en el aula para todos los estudiantes. La participación es voluntaria y la negativa a participar no implicará penalización ni pérdida de los beneficios a los que su estudiante tiene derecho. Para incluir aportes de su hijo/hija en este estudio, debemos obtener su consentimiento por escrito. Si está dispuesto a que su hijo participe en este proyecto de investigación, firme una copia del Formulario de consentimiento de los padres/tutores adjuntos de Weber State University. Firme y devuelva si acepta que su estudiante participe en el estudio. Gracias. Mis mejores deseos, Laura Wilcock Candidato a MEd y estudiante Weber State University 55 Appendix E: Items for object-based elicitation: 56 57