UNIVERSITY OF CALIFORNIA SAN DIEGO STATE UNIVERSITY “I’m a Biologist Who Codes”: An Examination of Computing Experiences and Persistence in Biology Along Educational Professionalization Trajectories A dissertation submitted partial satisfaction the requirements for degree Doctor Philosophy Mathematics Science Education by Austin Lee Zuckerman Committee charge: University California San Diego Professor Ashley Juavinett Chair Stanley Lo Co-Chair Mia Minnes State Gena Sbeglia Victoria Delaney 2025? Copyright L. Zuckerman 2025 All rights reserved The Dissertation is approved it acceptable quality form publication on microfilm electronically: California Table Contents Chapter 1. Introduction 1 Importance Computational Skills Thinking Biology 2 Opportunities to Learn K-12 Higher Education 5 Different Structures Introducing Biology 8 Professional Careers 9 Motivation Research Questions 11 2. Literature Review 14 Curricula other non-CS Disciplines 15 Contextualized programming courses Students 15 Attitudinal Barriers Toward Learning Computer Programming 18 Career Choices Persistence 21 Choice Pathways Computing 21 STEM (with an emphasis research paths) 25 Promoting Equity Pathways 28 Intentions Persist Undergraduate Education 29 3. Theoretical Conceptual Frameworks 34 Identity as Framing Participation Persistence 36 Conceptualizing Pathways 37 Examining Disciplinary Pathways 41 Framework Integrating Discipline-Based Theories 44 Integration Sociocultural Framework: Communities Practice 46 Features Practice 48 Multimembership Practice 52 Sociopolitical Rightful Presence 55 Operationalization Final Model 60 4. Methods 62 Phase 0: Pilot Study 68 Study Context 68 Findings 69 1: Programming Course Supporting Persist 74 Changes Curriculum 76 Assessment Survey 78 End Interviews 87 2: across Multiple Stages 87 Data Analysis 89 Quantitative Analysis Assessments Surveys (Phase 1) 89 Qualitative Interviews (Phases 1 & 2) 93 5. Validation Descriptive Statistics 98 Demographics respondent sample 100 Survey Constructs 107 Concept Inventory Assess Student Knowledge Programming 113 Relationships between Construct Measures 114 Measures 117 6. Outcomes Biology 119 Inferential statistics all survey items 120 Comparison outcomes course models 126 validated constructs 126 Linear mixed effects models 139 Measured Constructs with Significant Following Instruction 141 Differences construct measures student background variables models 164 Associated Persist 166 Profiles engagement introductory computer science courses 171 Latent profile analysis 171 7. Insights into Students 181 Decisions pursue biology-based course 182 Expectations about learning programming 183 Beliefs programmers programming 184 8. Factors Underlying Pathways 189 Social Influences 191 Scaffolded Resources Settings 194 Concrete Experiential Experiences 196 Intrinsically Motivating Informal Projects 199 Interdisciplinary experiences reinforce possibilities 201 9. Positioning Identities Across Practice 204 Stronger ties biology than computing (22 cases) 206 Fluid spaces (8 cases) 211 (4 cases) 213 Misfitting both communities (3 cases) 215 10. Identifying “Computational” Versus “Programmer” 217 What means be ‘computational’ 217 Focus area 218 Mathematical Affinity 221 computational biology 223 needed computational 225 ‘programmer’ 229 Distinguishing ‘Programmer’ from Other Identities 234 Influence sociocultural professional norms identity 237 Relative value skills 238 problem-solving approaches 242 Standards rigor prowess applications 244 11. Into How Recognition-based Formation 246 Context-Dependent Nature Recognition 248 Value Recognition Based Authority 251 Through Collaborative Mentorship Teaching Experiences 255 12. Perceived Biology 257 Barrier: Cumulative opportunity gaps due delayed exposure 259 Solution: Early exposure consistent applied opportunities curriculum 261 Varied effectiveness self-guided experiences 265 More scaffolds graduate education research 268 Marginalizing stereotypes spaces 271 Culture Brilliance 272 Fitting mold programmer 273 Marginalization through intersection race gender 276 Emotionally-responsive spaces 278 Lack alternative career paths academia industry 283 Expanded recognition structures non-traditional pathways computing 285 Incentive systems undermine democratization skills 286 Incentivization good practices open source contributions 290 13. Resisting Psychosocial Biology 292 Acknowledging identity asset work 293 empowering spaces 298 Being forced learn out necessity but guardrails 301 Reframing mindsets expertise engage spaces 302 14. Debate Theory Introductory Education 304 Need Education 305 Finding right balance abstraction theory application 307 Statistical-based introduction quantitative fundamentals 311 15. Discussion 315 Addressing Aim #1 316 Characteristics students who take contextualized biology 317 Belonging disciplinary communities 318 Interventions that address mindset identity 321 Limitations Future Work 321 #2 322 positioning practice 323 Engagement cycles 325 Affirming rightful presence within computing 328 Connecting Micro Macro Scale Insights 332 Inequities marginalization social identities 332 Comparisons traditional contemporary trajectories 334 Authenticity Applied Experiences 337 skills are recognized 339 Work 342 16. Recommendations Improving Biology 344 Reform undergraduate curriculum 344 Recommendation requirement majors 345 applications more ubiquitously curriculum 347 3: first or second year major 348 4: Scaffolding project-based curricula major 349 Inclusion curricular activities coding practices 351 aimed at improving individual development progression 352 Explicit conversations various contexts 353 Offering informal community partnerships events near peer students 354 Progressing culture skill laboratories 356 9: Creating curated resources code sharing allocated time workflow 356 10: collaborative growth-oriented culture 358 11: Building sustainable mentorship 359 Reimagining industry support expanded talent pool 361 12: Microcredential plausible pathway 361 13: Expanding new roles industry 362 14: pool instructors self-taught pathways 364 15: academic incentive system account contributions infrastructure 365 Emerging Considerations 366 Strategic sequencing courses 367 17: Reducing entry barriers explicit instruction responsible use generative AI 368 17. Concluding Remarks 370 References 373 Appendix A. Assessment 397 B. Attitudes Survey 400 C: #1 Interview Protocol 428 D: #2 Protocol 430 ? LIST FIGURES Figure 3.1: theories investigate contextual factors influence contexts. Integrated framework adapted conceptual model introduced Pfeifer et al. (2024). 46 3.2: practice lens navigation multiple communities. include multi-membership different potential trajectories practice. 48 3.3: sociopolitical framing model. systemic relations power equity shape formation communities. 58 3.4: used guide aims methodology interpretation findings. This study takes multilevel approach examining professionalization biologists stages novice expert continuum educational higher can facilitate biology. 61 4.1: Summary observations pilot focus groups framed cognitive (SCCT). Reproduced (2024). 70 4.2: Box whisker plot comparing pre- post-course fixed toward problem solving scale. White circle indicates mean values whiskers represent 25th 75th quartiles. P-values were calculated using independent t-test. **p<.01 72 5.1: Upset plots showing demographic sociodemographic each included subsequent analyses. 104 5.2: Bar graphs breakdown prior experience based major categories. three comparative analyses. 106 5.3: Wright maps Rasch-transformed person abilities (left) item difficulties (right) survey. post anchored pre-course estimation. 111 5.4: map abridged SCS1 assessment. estimation. 114 5.5: Heatmaps relationship Rasch- transformed constructs. 116 5.6: Heatmap differences assessment constructs. Only completed surveys points included. 118 6.1: Pre-and Rasch subscale. There no significant timepoints any model. 128 6.2: growth model. 129 6.3: self-efficacy One showed difference timepoints. *** p<0.001. 129 6.4: verbal persuasion models p<0.001. 130 6.5: interest model. 130 6.6: sense belonging p<0.001*p<0.05 131 6.7: intentions persist model. 131 6.8: gains knowledge measured SCS1. p<0.001. 132 6.9: Mean compatibility ‘Me’ ‘Computer Programming’ identities. Compatibility disaggregated administration timepoint. Error bars SEM. 136 6.10: ‘computer programming’ identities SEM. 137 6.11: following instruction. 148 6.12: instruction. 149 6.13: instruction. 154 6.14: instruction. 155 6.15: instruction. 156 6.16: instruction. 157 6.17: biology. 159 6.18: instruction. 161 6.19: instruction. 163 6.20: Relationship estimated predicted values every unit increase raw affective constructs. 169 6.21: Boxplots responses items ‘intentions persist’ construct Lower upper edges boxes percentiles median between. 170 6.22: Four profiles response patterns latent constructs results analysis. 173 6.23: Sankey Plot changes membership survey. 174 6.24: (indicated white marker center box). 180 15.1: Two taken when traversing practice. 325 15.2: work cycles identity. 328 15.3: practice 332 15.4: historical biology. 337 TABLES relevant concepts theoretical frameworks data collected question. 64 collection timeline. 67 4.3: administered participating courses. For varying number items additional contexts course. 86 Response rate assessment course. 100 respondents model. 102 fit indices CFA analyses Cronbach’s alpha internal reliability subscale. 108 Item separation reliabilities eigenvalue contrast residuals attitudinal survey. 110 assessment. 113 courses. 121 Effect sizes (as Cohen’s d) have least one pre measures. 133 inferential perceived subgroups biology. 138 regression constructs. 143 measures generated part multivariate including outcome variables. 146 variables. 147 self-efficacy persuasion interest variables. 150 variables. 151 variables. 152 variables. 153 separate only includes biology. 158 variable. 160 variable. 162 variable (using values). 168 sample. 175 sample. 177 chi-square tests timepoint variables 178 Results linear predicting controlling time. 181 7.1: Background characteristics interviewed discipline-based biology. 182 8.1: Description participants full interview (n=37). 190 8.2: themes influential aspects identified category. 191 9.1: Outcome space variations four conceptions practice. 205 10.1: areas differentiated how their identity. 218 10.2: ways aligned mathematical affinity identity. 221 10.3: attributes described necessary claiming ‘computational.’ 226 10.4: descriptions ‘programmer’. 230 10.5: Frequency coded memberships labels “computational person” “programmer” participants. 235 10.6: describing navigation. 238 11.1: recognition-based computing. 248 12.1: navigating solutions proposed those barriers. 258 13.1: Categories assets they could leverage communities. 293 ACKNOWLEDGMENTS I begin immense amount debt gratitude incredible mentors I’ve had throughout this process. feel extremely privileged struck meaningful autonomy nurture my researcher. To Dr. thank you being visionary pioneering me. remember taking little did know we would end up working so closely together projects become some most studies. Your active listening ability intuit act upon needs myself others steer bigger picture has offered much grounding which definitely constant reassurance encouragement not go unnoticed. deeply appreciated available you’ve always been offer advice gut check just listening ear. That kind makes real difference. I’m constantly inspired your tireless dedication advocacy commitment reforming education look forward continuing grow our its next evolution. also want extend Yeti his emotional reminding us team four-legged co-PI. mentor me since was undergraduate. incredibly grateful shepherded DBER remained trusted ever since. lost count we’ve worked over years privilege experience. early belief gave confidence might otherwise imagined. You continue brilliant force nature pushing forward am continually awe while maintaining such deep field. As move stage carry everything taught me continued you. Thank everything. Sbeglia caring approachable easy talk to. learned done together hope standards measurement further DBER. Every met you left renewed direction clearer vision impactful researcher communicator. You’ve generous insights provided work. remarkable make complex ideas accessible meaningful created positive welcoming environment group. think critically actionable. Miranda Parker Minnes Delaney input committee members. extended welcome reached nothing gracious feedback unique thoughtful guidance invaluable truly appreciate effort invested helping refine research. like MSED community. Deb Escamilla Sherry Seethaler lifesavers fielding incessant questions advocate program. Deb heart CRMSE — kindness support calming energy made world program same without Susan Nickerson supportive resource. played crucial role making whirlwind joint doctoral manageable. fight goals lasting impact helped create stronger us. Donna Ross bright light openness difference joy see Cadi office. warmth positivity contagious shared along way. Nicole Suarez crux during mentorship after graduation. affirmation moral support. Michelle Nolasco immediately offering entered regular chats share office Wembley. generosity. sharing together. shy ideas visions. super inspiring visionaries aiming mountains education. Never lose spark imagine future push boundaries what be. outside profoundly impact school Valentin Cracan Xingxiu Pan Mina Heacock unconventional giving contribute lab way challenged lessons gained will stay educator. Prashant Mantha Jocelyn Newsome Anisa Abeytia invaluable mentorship. Working tremendously expansively kinds ask serve. These excited colleagues mentees sectors impactful. And finally risk forgetting names heartfelt thanks friends family there process – are. encouragement patience laughs academia. endlessly shown comes next. ABSTRACT THE DISSERTATION Diego University M. increasingly recognized rather supplemental competency However many life sciences programs lack adequate coursework understanding navigate remains limited. financial implications training mandate innovative strategies invite discipline-relevant equitable. broad purpose explore underlie participation persistence bridge biology classroom postgraduate experiences. adopts integrated combines (community practice) (rightful presence) lenses examine individual contextual biologists’ pathways. complementary aims cross-sectional case introducing biology. modeling compare students' types courses well disparities groups. analysis then identify distinct these outcomes. findings promising compromising key examines international pathways students postdocs university faculty professionals. Findings phenomenographic reflexive thematic culminate recommendations computing addressing challenges supporting labs reimagining versatile careers. 1. Chapter Introduction rapid evolution continues revolutionize society. In writing increased visibility large language artificial intelligence begun excitement curiosity disciplines among general public. breakthroughs intersected mainstream regarding importance achieving 21st century. citizens professions especially rapidly growing personal technology advancement global economy unprecedented (Bocconi al. 2016 Mohaghegh McCauley 2016). consider overwhelmingly shaped defining operationalizing term ‘computational’ straightforward feat. often situated ‘computational thinking’ disciplines. thinking construed thought where logic solve problems surrounding (Wing 2006). Individuals proficient type able “deconstruct abstract generalize information sequentially algorithmically explain phenomena” (Peel 2021 p. 112). regarded integral digital world thereby increasing fundamental literacy restricted scientists (Gouvea 2023 Wing definitions emphasize application mathematics definition literacy’ material cognitive (diSessa 2000 Jacob Warschauer 2018 Odden 2019). formalized (2019) pillars read write manipulate (material) capitalize tools (cognitive) artifacts communicate others (social). Although scholars argue (Bell 2009) central operationalized. Especially scientific disciplines emphasized consists related simulation thinking solving expression (Arastoopour Irgens 2020 Weintrop everyone programmers developing similar transferable range fields. Scientific domains reliant technologies drive innovation (Bundy 2007). Among science technology engineering mathematics (STEM) fields hold substantial public health energy environmental regulation sectors. Consequently past few decades field biological capacities high throughput innovations evidenced discourse ‘big data’ (Schatz 2012 Yin 2017). particular bioinformatics subfields bridged applying methods applications -omics era (Markowetz Given ubiquity because requisite set thrive today’s workforce mastery analytics tend earn salaries participate greater breadth cutting edge (NCES Shah 2022 Way 2020). salary mobility raise accessibility acquire sets demand traction fields science. even expanding variety example gender racial documented Advanced Placement classes (Ericson Guzdial 2014). debts rates females racial/ethnic minority discipline level stem largely socioeconomic inequities reduce access immersion extracurricular (Tsan Wang Hejazi Moghadam differential reproduce inequalities (Margolis 2008). sustained insufficient representation prevail progress workforce. majors lines despite anomalies (e.g. Black master’s level) (NCSES 2023). reported 2021 Earned Doctorates (SED) National Center Engineering Statistics (NCSES) 25% 2022 recipients female less 20% non-Asian non-White 2022). relative improvements recruiting women color decade progression parity slow expand (Lehman contrast experienced substantially 55% female interdisciplinary important acknowledge may observed uniformly subfields intersect engineering subdisciplines underrepresented interdisciplinary 32% provide inviting yet meta-analysis demonstrated lower authors publications (Bonham Stefan Racial year ten Hispanic fewer individuals bioinformatics compared hundreds biomedical While reasons underrepresentation multifaceted complex psychosocial motivation prevailing people uninspiring unwelcoming common stereotype antisocial males competitive personalities (Cheryan 2009 Lewis perceptions adopt essentialist undermines (Berg Cheryan 2013 2015 Kendall 2011 Spieler Wong computationally intensive compromise if fail implement strategic interventions resist marginalizing narratives. necessitates avenues combating exclusion (Adams barrier progressing United States standardized curriculum comparably non (non-CS) Initiatives scale particularly (Smith Ryoo report national Code.org half schools teach concepts leave void (Code.org CSTA ECEP Alliance Another found 75% principals programming home lines (Wang challenging inadequate teacher coherent (Goode Vogel philosophy argued curricula reformers arguing do ‘own’ (Dodds 2021). Many universities decade need establish cohesion experimental (Noble von Arnim Missra Despite century indispensable advancing institutions specialization (Pevzner Shamir 2009). neuroscience programs 10 118 required and/or elective (Pinard-Welyczko 2017 cited It 15% electives indicating insufficiently (Society Neuroscience Current meet tasks data calls institutions funding agencies supports gap (Barone Proposed lead (1) sufficient pedagogical skills (2) languages environments (3) content relatable interesting (Juavinett Williams teaching generally valued coding modeling reproducibility visualization (Emery collectively interested integrate direct investigation oriented place. students’ evaluate enjoyment (Kapoor Gardner-McCune examples basic trend (LeBlanc Dyer 2004 Dodds 2012). prerequisites requirements. lieu formal course alternatively bootcamps gain (Thayer Ko impose burdens constraints guarantee employers recognize legitimate substitute training. Emphasizing forms literacies issues cognition learning justice Kafai Proctor (2021) put forth considerations integrating computation environments: whom currently designed for counts practice should taught. initially conceptualized whether taught implicit signals Even reform initiatives authentic inquiry (AAAS 2011 Freeman 2014 Theobald 2020) textbooks promote rote memorization thus depriving insights (Buxton Feser 2013). still efforts understand best introduce students. Careers boundary blurred assessed utility specifically (Shah reinforced settings reporting important. Importantly professionals issue equity given mobility. elicits question engaging actually acquired training opportunities advanced positions. graduated pursued careers enhance researchers online websites Kaggle Carpenteries DataCamp workshops boot camps clubs (Hagan curation resources suggested professionals rely self-learning collaborations skills. viable self-motivated robust infrastructure ensure (Riquelme Gjorgjieva Furthermore investigating useful existing face allowed them is current focuses improved equally keep limited involvement focus. elucidate initial participation beyond leveraging understudied population fill niche optimal They considerably successful identifying critical Drawing networks instrumental persevering development felt beneficial earlier trajectories. zoomed examination informative target inclusion Questions pursuing context multiscale stems forged. First scale achieved exploring curricula. Second deeper experiences document perspectives computing. At extreme embarking committed intermediate laboratories post-baccalaureate Exploring potentially uncover pivotal decision choices Because portion better meaningfully pairing broader ‘reverse engineer’ promoting considering represents point decisions made progressed later Mapping pinpoint junctions interventions systems strategically deployed inform ongoing generation biologists. intended accounting scales holistic factors divided two (italicized) corresponding questions. corresponds structure #1: Investigating (a class biology). 1.1. extent does affect programming? 1.2. learning? 1.3. computing? 1.4. courses? 1.5. entering #2: novice-expert continuum. 2.1. biology? 2.2. position identities spaces? 2.3. pathways? 2. Chapter Review chapters consist literature review overview frameworks guides rationale methodology aims. review situate relevance (Lufts establishing interpret (Bussey integrates describes relationships main investigation. guided framing typically encompassing emerging Before parallels lays groundwork approaches pre-existing reviewed. known strong base examined broadly. influencing literature several methodologies Disciplines Students seeks efficacy beliefs section Within debatable modules educators laboratory organically hands-on data. tutorials statistical (Custer 2021) (Madlung 2018) desire comparable CS-1 prerequisite US Academies approximately one-third enrolled majoring (NASEM 2018). practitioners diverse backgrounds Pruim Camp 2017 Dawson Sax come ‘one size fits all’ apply CS futile reduced result (Camp Hogan dropout (Robins 2003 Watson Li Yadin 2011). avenue providing non-majors (e.g. Guzdial Forte 2005 Urban-Lurain Weinshank 1999). Benefits increases (Hogan 2023) (Dawson non-major Courses under themed contextualized deem