About

Rachel B. Baker is an associate professor in the Policy, Organizations, Leadership, and Systems Division at the Graduate School of Education at the University of Pennsylvania. Her research focuses on issues related to access to and success in higher education, with particular attention to students in broad-access institutions. She studies how various policies influence community college students' decisions about majors and transfer, describes enrollment patterns by race and socio-economic status over time, and measures students’ self-regulatory skills using interaction data from online classes. Baker is involved in a five-year NSF-funded study examining the effects of cross-enrollment on academic aspirations and transfer success, and she leads efforts to develop measures of curricular complexity in college major requirements. Her work aims to increase access, persistence, and success for traditionally underserved groups in higher education.

Research topics

• Computer Science
• Sociology
• Political Science
• Data science
• Psychology
• Social psychology
• Machine Learning
• Artificial Intelligence
• Data Mining
• Epistemology
• Engineering ethics
• Human–computer interaction
• Engineering
• World Wide Web
• Mathematics education

Selected publications

• Perceptions of concussion management and testing among healthcare professionals in rugby union: a qualitative analysis

  BMJ Open Sport & Exercise Medicine · 2026-04-01

  articleOpen access

  Despite ongoing developments in head injury protocols, rugby union still faces challenges in achieving accurate, rapid and context-appropriate concussion assessment on-field and at pitch-side, with clinicians often required to make complex decisions using tools that remain imperfect. The purpose of this research is to gain insight into the perceptions and experiences of high-level rugby union healthcare professionals (HCPs) on current head injury assessment protocols. Semi-structured interviews were conducted with eight HCPs working in rugby union. Following a six-stage, inductive, reflexive thematic analysis, five key themes emerged. Findings reveal that while standardised protocols offer structure to an assessment, clinical judgement remains central to effective diagnosis. Participants expressed cautious optimism towards emerging technologies, including biomarker testing and motor control assessments. This study offers practitioner-informed insights into the nuanced realities of concussion assessment. Furthermore, this work underscores the need for future tools to be evidence-based and adaptable to the real-world context of elite sport.

  PublisherOA PDFDOI

• Practice as the Key to Success: Understanding the Role of Prior Knowledge, Affective States and Learning Resources in Computer Science Education

  2026-04-25

  articleOpen accessSenior author

  Introductory programming courses (CS1) bring together students with diverse prior experiences, which shape their use of learning resources, emotional responses, and academic performance. This study employs structural equation modeling, self-reported affective data, and multimodal interaction logs to investigate how prior programming knowledge affects affect, resource utilization, and outcomes in a CS1 course that features an automated assessment tool (AAT), instructional videos, and worked examples. Students who persisted with practice and advanced beyond basic tasks achieved the strongest outcomes, though they followed different emotional pathways. By contrast, relying solely on videos or worked examples did not significantly lead to success, and disengagement was generally tied to weaker performance. Novices often reported confusion and frustration; while these emotions sometimes hindered learning, they also drove deeper engagement with the AAT, improving outcomes for those who persisted. Experienced students, however, more often reported boredom, which consistently reduced practice and led to poorer outcomes. These findings underscore the need for adaptive support that balances challenge with guidance to sustain engagement and promote success for diverse learners in CS1.

  PublisherDOI

• Capturing Professional Skill Development: A Curriculum Analytics Approach

  2026-04-25

  articleOpen access

  Higher education faces increasing pressure from governments and employers to ensure graduates are equipped with the knowledge and capabilities required for the future workplace. While technical knowledge is assessed through grades, professional skills are complex and remain difficult to evaluate systematically. While curriculum mapping offers a potential solution, it is often applied in a simplistic, accreditation-driven manner that merely records the presence or absence of skills embedded in assessments. Such an approach overlooks the relative contribution of each skill to the assessments and, hence, cannot be used to estimate skill development. Other noted approaches have relied on the use of self-assessment reports or surveys, and as such are subjective and cannot provide longitudinal evidence of skill development. This study addresses these limitations by proposing a novel curriculum analytics method, weighted Performance Factor Analysis, to model skill scores using assessment grades and granular weighted skill-assessment mapping. Students’ development of seven professional skills were examined along with how they transition across an accounting degree program. The findings show distinct patterns and trajectories of skill development. Overall, the study makes a significant methodological contribution to measuring skills and offers insights into how graduates develop their professional skills across the curriculum.

  PublisherDOI

• Do MOOC Conversations Matter? Investigating the Role of Social Presence and Course-Relevant Discussion in Career Advancement

  2025-07-17

  articleSenior author

  While MOOCs have been widely studied in terms of student engagement and academic performance, the extent to which engagement within MOOCs predict career advancement remains underexplored. Building on prior work, this study investigates how participation in discussion forums, specifically social presence and the use of course-relevant keywords, affects career advancement. Using GPT-assisted content analysis of forum posts, we assess how these engagement factors relate to both achievement during the course and post-course career advancement. Our findings indicate that social presence and use of course-relevant keywords has a positive relationship with course achievement during the MOOC. However, no significant relationship was found between career advancement and either social presence or course-related keywords in discussion forums. These findings suggest that while active engagement in MOOC discussion forums enhances academic achievement, it might not directly translate into career advancement, highlighting a possible disconnect between learning participation in MOOCs and professional outcomes.

  PublisherDOI

• Data Plus Theory Equals Codebook: Leveraging LLMs for Human-AI Codebook Development

  2025-09-28

  articleOpen access

  Recent research has explored the use of Large Language Models (LLMs) to develop qualitative codebooks, mainly for inductive work with large datasets, where manual review is impractical. Although these efforts show promise, they often neglect the theoretical grounding essential to many types of qualitative analysis. This paper investigates the potential of GPT-4o to support theory-informed codebook development across two educational contexts. In the first study, we employ a three-step approach—drawing on Winne & Hadwin’s and Zimmerman’s Self-Regulated Learning (SRL) theories, think-aloud data, and human refinement—to evaluate GPT-4o’s ability to generate high-quality, theory-aligned codebooks. Results indicate that GPT-4o can effectively leverage its knowledge base to identify SRL constructs reflected in student problem-solving behavior. In the second study, we extend this approach to a STEM game-based learning context guided by Hidi & Renninger’s four-phase model of Interest Development. We compare four prompting strategies: No theories provided, theories named, full references given, and full-text theory papers supplied. Human evaluations show that naming the theory without including full references produced the most practical and usable codebook, while supplying full papers to the prompt enhanced theoretical alignment but reduced applicability. These findings suggest that GPT-4o can be a valuable partner in theory-driven qualitative research when grounded in well-established frameworks, but that attention to prompt design is required. Our results show that widely available foundation models—trained on large-scale open web and licensed datasets—can effectively distill established educational theories to support qualitative research and codebook development.

  PublisherDOI

• Exploring Differences Between Hybrid GPT-Human and Human-Created Qualitative Codebooks in an Educational Game

  Communications in computer and information science · 2025-12-01

  book-chapter
  PublisherDOI

• On-pitch concussion management: a view of healthcare professionals in elite football

  BMJ Open Sport & Exercise Medicine · 2025-10-01

  articleOpen access1st authorCorresponding

  Collecting an account of practitioners' lived experiences of on-pitch concussion management in football provides real-world insight into areas for improvement, and future research direction. Previous research suggests that healthcare professionals (HCPs) in football are confident with on-pitch concussion recognition; however, some players are still returned to the field to later be removed with a concussion. Objectives: The aim of this study is to understand the on-pitch assessment procedure of HCPs and factors that may influence whether or not a player is removed following a suspected head injury in football based on semistructured interviews. Methods: 10 elite-level pitch-side HCPs participated in a semistructured interview. Recorded interview transcripts were analysed via reflexive thematic analysis. This process generated four key themes: time, formalised procedure for identifying concussion, interpretation of behaviour and stakeholder actions. Results: This study provides evidence that HCPs in football feel that they do not have enough time to conduct a thorough concussion assessment. Assessments are currently subjective and rely on the sideline healthcare staff's relationship with the player to determine if they are acting normally. The HCPs value seeing the mechanism of injury and advocate the use of video replays. Temporary concussion substitutes would allow more time for assessment and for the use of assessment tools such as the Sport Concussion Assessment Tool. Conclusions: Based on the current findings, research should focus on developing multimodal testing batteries to detect concussion on-pitch, while football should consider implementing video analysis at all levels and temporary concussion substitutes.

  PublisherDOI

• Downshifting: The Nuanced Antecedents of Seeking Easier Content

  Proceedings. · 2025-06-10

  articleOpen access

  Learning involves challenge, and students respond to challenging content in different ways.This study investigates the antecedents and contexts of downshifting: seeking out easier questions while using a digital learning platform.We study this behavior using student data from Rori -a chatbot-based math intelligent tutoring system that operates over WhatsApp and is widely used in several African countries.Specifically, we use Sequential Pattern Mining to study the behaviors that immediately precede and follow downshifting and Bayesian Knowledge Tracing to evaluate students' mastery of the content prior to downshifting.Results show that the behavior is often preceded by sequences of incorrect answers, suggesting that some students may be adaptively downshifting to avoid failure.However, some students seem to downshift despite being on track to master the current content.We also find that students using Rori in schools as part of their regular curricula are likelier to downshift than those who access Rori independently.We position downshifting within a general taxonomy of behaviors related to persistence in ITSs and suggest further research into the causes of the behavior.

  PublisherOA PDFDOI

• A Step Towards Adaptive Online Learning: Exploring the Role of GPT as Virtual Teaching Assistants in Online Education

  2025-10-10

  book-chapterSenior author

  With student learning becoming more continuous and ubiquitous, online courses are increasingly challenged to provide timely support to learners. Human TAs, constrained by limited availability, often delay in addressing student inquiries occurring on weekends or at night. This paper presents JeepyTA, a Virtual Teaching Assistant (VTA) built on GPT model designed to provide round-the-clock assistance by leveraging OpenAI’s text embeddings and generative language models. JeepyTA provides responses that mimic typical discourse in discussion forums and, although still limited in scope, addresses logistic, conceptual, and programming questions tailored to specific courses much quicker than human TAs can. In this paper, we outline our development process, analyze JeepyTA’s response accuracy and compare its availability to human instructors, investigate student attitudes, and discuss the implications of integrating large language models like JeepyTA into educational settings. This work contributes to understanding how LLMs could improve the timeliness and availability of student support, offering on-the-spot assistance, and delivering personalized feedback.

  PublisherDOI

• The effect of football (soccer) heading on gross and fine motor control in women

  Frontiers in Sports and Active Living · 2025-07-02

  articleOpen access

  Introduction: Heading is an integral component of football, but concerns remain about its impact on brain health. This study examines the acute effects of heading on gross and fine motor control as a measure of the motor-cognitive function of women footballers. Methods: The heading protocol for this study represented the typical exposure to headers experienced in the women's game: one every 10 min, for one hour, replicating a corner kick. A sample of 19 female collegiate football (soccer) players participated in two sessions: a control session, and a heading intervention. Gross motor control was assessed via measures of sway during standing balance, and fine motor control was evaluated using a precision finger grip task. Results: Results showed no significant changes in gross motor control, based on postural sway parameters. However, significant alterations were observed in fine motor control in the tremor frequency (0-4 Hz band) of precision gripping, indicating a potential change in motor-cognitive function following the heading task. Discussion: The findings suggest that exposure to the number and type of headers that might be performed over a typical football match does not impair standing balance, but it may affect fine motor control. Future research should look to incorporate brain imaging and electrophysiological measures to further understand the mechanisms underpinning changes in fine motor control performance after heading.

  PublisherOA PDFDOI

Recent grants

• Collaborative Research: The Downside of Perseverance--Investigating and Moving Students Beyond Unproductive Persistence

  NSF · $400k · 2015–2020

• Collaborative Research: Using Data Mining and Observation to derive an enhanced theory of SRL in Science learning environments

  NSF · $862k · 2016–2017

• Collaborative Research: Frameworks: Cyber Infrastructure for Shared Algorithmic and Experimental Research in Online Learning

  NSF · $1.4M · 2019–2025

• Collaborative Research: Student Affect Detection and Intervention with Teachers in the Loop

  NSF · $236k · 2019–2024

• Collaborative Research: Using Educational Data Mining Techniques to Uncover How and Why Students Learn from Erroneous Examples

  NSF · $585k · 2017–2022

Frequent coauthors

• Jaclyn Ocumpaugh

  University of Pennsylvania

  111 shared

• Luc Paquette

  48 shared

• Shamya Karumbaiah

  Carnegie Mellon University

  44 shared

• Neil T. Heffernan

  Worcester Polytechnic Institute

  43 shared

• Albert T. Corbett

  40 shared

• Stephen Hutt

  University of Denver

  37 shared

• Kenneth R. Koedinger

  Carnegie Mellon University

  35 shared

• Ma. Mercedes T. Rodrigo

  Ateneo de Manila University

  30 shared

Awards & honors

• 2019 NAEd/Spencer Foundation Postdoctoral Fellow
• 2016 Outstanding Fellow from the Institute of Education Scie…
• Jack Kent Cooke Foundation Dissertation Fellowship