Conrad Borchers

Liang Zhang, Jionghao Lin, John Sabatini et al. · cmu

Learning performance data describe correct and incorrect answers or problem-solving attempts in adaptive learning, such as in intelligent tutoring systems (ITSs). Learning performance data tend to be highly sparse (80\%\(\sim\)90\% missing observations) in most real-world applications due to adaptive item selection. This data sparsity presents challenges to using learner models to effectively predict future performance explore new hypotheses about learning. This article proposes a systematic framework for augmenting learner data to address data sparsity in learning performance data. First, learning performance is represented as a three-dimensional tensor of learners' questions, answers, and attempts, capturing longitudinal knowledge states during learning. Second, a tensor factorization method is used to impute missing values in sparse tensors of collected learner data, thereby grounding the imputation on knowledge tracing tasks that predict missing performance values based on real observations. Third, a module for generating patterns of learning is used. This study contrasts two forms of generative Artificial Intelligence (AI), including Generative Adversarial Networks (GANs) and Generate Pre-Trained Transformers (GPT) to generate data associated with different clusters of learner data. We tested this approach on an adult literacy dataset from AutoTutor lessons developed for Adult Reading Comprehension (ARC). We found that: (1) tensor factorization improved the performance in tracing and predicting knowledge mastery compared with other knowledge tracing techniques without data augmentation, showing higher relative fidelity for this imputation method, and (2) the GAN-based simulation showed greater overall stability and less statistical bias based on a divergence evaluation with varying simulation sample sizes compared to GPT.

Conrad Borchers, Dalia Sara Gala, Benjamin Gilburt et al. · oxford

The growing capability and availability of generative language models has enabled a wide range of new downstream tasks. Academic research has identified, quantified and mitigated biases present in language models but is rarely tailored to downstream tasks where wider impact on individuals and society can be felt. In this work, we leverage one popular generative language model, GPT-3, with the goal of writing unbiased and realistic job advertisements. We first assess the bias and realism of zero-shot generated advertisements and compare them to real-world advertisements. We then evaluate prompt-engineering and fine-tuning as debiasing methods. We find that prompt-engineering with diversity-encouraging prompts gives no significant improvement to bias, nor realism. Conversely, fine-tuning, especially on unbiased real advertisements, can improve realism and reduce bias.

Conrad Borchers, Zachary A. Pardos

Course load analytics (CLA) inferred from LMS and enrollment features can offer a more accurate representation of course workload to students than credit hours and potentially aid in their course selection decisions. In this study, we produce and evaluate the first machine-learned predictions of student course load ratings and generalize our model to the full 10,000 course catalog of a large public university. We then retrospectively analyze longitudinal differences in the semester load of student course selections throughout their degree. CLA by semester shows that a student's first semester at the university is among their highest load semesters, as opposed to a credit hour-based analysis, which would indicate it is among their lowest. Investigating what role predicted course load may play in program retention, we find that students who maintain a semester load that is low as measured by credit hours but high as measured by CLA are more likely to leave their program of study. This discrepancy in course load is particularly pertinent in STEM and associated with high prerequisite courses. Our findings have implications for academic advising, institutional handling of the freshman experience, and student-facing analytics to help students better plan, anticipate, and prepare for their selected courses.

Yishan Du, Conrad Borchers, Mutlu Cukurova

As teachers increasingly turn to GenAI in their educational practice, we need robust methods to benchmark large language models (LLMs) for pedagogical purposes. This article presents an embedding-based benchmarking framework to detect bias in LLMs in the context of formative feedback. Using 600 authentic student essays from the AES 2.0 corpus, we constructed controlled counterfactuals along two dimensions: (i) implicit cues via lexicon-based swaps of gendered terms within essays, and (ii) explicit cues via gendered author background in the prompt. We investigated six representative LLMs (i.e. GPT-5 mini, GPT-4o mini, DeepSeek-R1, DeepSeek-R1-Qwen, Gemini 2.5 Pro, Llama-3-8B). We first quantified the response divergence with cosine and Euclidean distances over sentence embeddings, then assessed significance via permutation tests, and finally, visualised structure using dimensionality reduction. In all models, implicit manipulations reliably induced larger semantic shifts for male-female counterfactuals than for female-male. Only the GPT and Llama models showed sensitivity to explicit gender cues. These findings show that even state-of-the-art LLMs exhibit asymmetric semantic responses to gender substitutions, suggesting persistent gender biases in feedback they provide learners. Qualitative analyses further revealed consistent linguistic differences (e.g., more autonomy-supportive feedback under male cues vs. more controlling feedback under female cues). We discuss implications for fairness auditing of pedagogical GenAI, propose reporting standards for counterfactual evaluation in learning analytics, and outline practical guidance for prompt design and deployment to safeguard equitable feedback.

Conrad Borchers, Robin Schmucker, Ashutosh Tiwari et al.

Cross-enrollment across institutions can expand access to courses and support student progression. Still, little is known about how geographic constraints and institutional policies jointly shape cross-enrollment within community college (CC) systems. We adopt a push-pull framework: geographic remoteness constrains feasible cross-institution mobility, while credit mobility may attract enrollment expressed as articulation (CC-to-university: credit toward a four-year partner) and course equivalencies (CC-to-CC: equivalencies across the system). Using de-identified administrative records from a 12-institution community college system (100,547 students; 1,290,311 course enrollments), we quantify outgoing and incoming cross-enrollment and relate these patterns to institutional remoteness and credit mobility. We find that less remote colleges exhibit higher outgoing and incoming cross-enrollment than more remote colleges. Further, cross-enrolled students are more likely to take articulated courses, and institutions with higher equivalency ratios receive higher incoming cross-enrollment (8.62% vs. 6.70%). This association was slightly stronger at more remote colleges. This study demonstrates how analysis of complex college systems can surface factors shaping student mobility and inform the design of cross-enrollment and articulation policies in CC systems.

Devika Venugopalan, Ziwen Yan, Conrad Borchers et al. · cmu

Caregivers (i.e., parents and members of a child's caring community) are underappreciated stakeholders in learning analytics. Although caregiver involvement can enhance student academic outcomes, many obstacles hinder involvement, most notably knowledge gaps with respect to modern school curricula. An emerging topic of interest in learning analytics is hybrid tutoring, which includes instructional and motivational support. Caregivers assert similar roles in homework, yet it is unknown how learning analytics can support them. Our past work with caregivers suggested that conversational support is a promising method of providing caregivers with the guidance needed to effectively support student learning. We developed a system that provides instructional support to caregivers through conversational recommendations generated by a Large Language Model (LLM). Addressing known instructional limitations of LLMs, we use instructional intelligence from tutoring systems while conducting prompt engineering experiments with the open-source Llama 3 LLM. This LLM generated message recommendations for caregivers supporting their child's math practice via chat. Few-shot prompting and combining real-time problem-solving context from tutoring systems with examples of tutoring practices yielded desirable message recommendations. These recommendations were evaluated with ten middle school caregivers, who valued recommendations facilitating content-level support and student metacognition through self-explanation. We contribute insights into how tutoring systems can best be merged with LLMs to support hybrid tutoring settings through conversational assistance, facilitating effective caregiver involvement in tutoring systems.

Conrad Borchers, Tianze Shou

Large Language Models (LLMs) hold promise as dynamic instructional aids. Yet, it remains unclear whether LLMs can replicate the adaptivity of intelligent tutoring systems (ITS)--where student knowledge and pedagogical strategies are explicitly modeled. We propose a prompt variation framework to assess LLM-generated instructional moves' adaptivity and pedagogical soundness across 75 real-world tutoring scenarios from an ITS. We systematically remove key context components (e.g., student errors and knowledge components) from prompts to create variations of each scenario. Three representative LLMs (Llama3-8B, Llama3-70B, and GPT-4o) generate 1,350 instructional moves. We use text embeddings and randomization tests to measure how the omission of each context feature impacts the LLMs' outputs (adaptivity) and a validated tutor-training classifier to evaluate response quality (pedagogical soundness). Surprisingly, even the best-performing model only marginally mimics the adaptivity of ITS. Specifically, Llama3-70B demonstrates statistically significant adaptivity to student errors. Although Llama3-8B's recommendations receive higher pedagogical soundness scores than the other models, it struggles with instruction-following behaviors, including output formatting. By contrast, GPT-4o reliably adheres to instructions but tends to provide overly direct feedback that diverges from effective tutoring, prompting learners with open-ended questions to gauge knowledge. Given these results, we discuss how current LLM-based tutoring is unlikely to produce learning benefits rivaling known-to-be-effective ITS tutoring. Through our open-source benchmarking code, we contribute a reproducible method for evaluating LLMs' instructional adaptivity and fidelity.

Conrad Borchers, Manit Patel, Seiyon M. Lee et al.

Open-ended responses are central to learning, yet automated scoring often conflates what students wrote with how teachers grade. We present an analytics-first framework that separates content signals from rater tendencies, making judgments visible and auditable via analytics. Using de-identified ASSISTments mathematics responses, we model teacher histories as dynamic priors and represent text with sentence embeddings. We apply centroid normalization and response-problem embedding differences, and explicitly model teacher effects with priors to reduce problem- and teacher-related confounds. Temporally-validated linear models quantify the contributions of each signal, and model disagreements surface observations for qualitative inspection. Results show that teacher priors heavily influence grade predictions; the strongest results arise when priors are combined with content embeddings (AUC~0.815), while content-only models remain above chance but substantially weaker (AUC~0.626). Adjusting for rater effects sharpens the selection of features derived from content representations, retaining more informative embedding dimensions and revealing cases where semantic evidence supports understanding as opposed to surface-level differences in how students respond. The contribution presents a practical pipeline that transforms embeddings from mere features into learning analytics for reflection, enabling teachers and researchers to examine where grading practices align (or conflict) with evidence of student reasoning and learning.

Zifei FeiFei Han, Jionghao Lin, Ashish Gurung et al. · cmu

One-on-one tutoring is an effective instructional method for enhancing learning, yet its efficacy hinges on tutor competencies. Novice math tutors often prioritize content-specific guidance, neglecting aspects such as social-emotional learning. Social-emotional learning promotes equity and inclusion and nurturing relationships with students, which is crucial for holistic student development. Assessing the competencies of tutors accurately and efficiently can drive the development of tailored tutor training programs. However, evaluating novice tutor ability during real-time tutoring remains challenging as it typically requires experts-in-the-loop. To address this challenge, this preliminary study aims to harness Generative Pre-trained Transformers (GPT), such as GPT-3.5 and GPT-4 models, to automatically assess tutors' ability of using social-emotional tutoring strategies. Moreover, this study also reports on the financial dimensions and considerations of employing these models in real-time and at scale for automated assessment. The current study examined four prompting strategies: two basic Zero-shot prompt strategies, Tree of Thought prompt, and Retrieval-Augmented Generator (RAG) based prompt. The results indicate that the RAG prompt demonstrated more accurate performance (assessed by the level of hallucination and correctness in the generated assessment texts) and lower financial costs than the other strategies evaluated. These findings inform the development of personalized tutor training interventions to enhance the the educational effectiveness of tutored learning.

Conrad Borchers, Jiayi Zhang, Ryan S. Baker et al.

Numerous studies demonstrate the importance of self-regulation during learning by problem-solving. Recent work in learning analytics has largely examined students' use of SRL concerning overall learning gains. Limited research has related SRL to in-the-moment performance differences among learners. The present study investigates SRL behaviors in relationship to learners' moment-by-moment performance while working with intelligent tutoring systems for stoichiometry chemistry. We demonstrate the feasibility of labeling SRL behaviors based on AI-generated think-aloud transcripts, identifying the presence or absence of four SRL categories (processing information, planning, enacting, and realizing errors) in each utterance. Using the SRL codes, we conducted regression analyses to examine how the use of SRL in terms of presence, frequency, cyclical characteristics, and recency relate to student performance on subsequent steps in multi-step problems. A model considering students' SRL cycle characteristics outperformed a model only using in-the-moment SRL assessment. In line with theoretical predictions, students' actions during earlier, process-heavy stages of SRL cycles exhibited lower moment-by-moment correctness during problem-solving than later SRL cycle stages. We discuss system re-design opportunities to add SRL support during stages of processing and paths forward for using machine learning to speed research depending on the assessment of SRL based on transcription of think-aloud data.

Liang Zhang, Jionghao Lin, Conrad Borchers et al. · cmu

Intelligent Tutoring Systems (ITSs) have significantly enhanced adult literacy training, a key factor for societal participation, employment opportunities, and lifelong learning. Our study investigates the application of advanced AI models, including Large Language Models (LLMs) like GPT-4, for predicting learning performance in adult literacy programs in ITSs. This research is motivated by the potential of LLMs to predict learning performance based on its inherent reasoning and computational capabilities. By using reading comprehension datasets from the ITS, AutoTutor, we evaluate the predictive capabilities of GPT-4 versus traditional machine learning methods in predicting learning performance through five-fold cross-validation techniques. Our findings show that the GPT-4 presents the competitive predictive abilities with traditional machine learning methods such as Bayesian Knowledge Tracing, Performance Factor Analysis, Sparse Factor Analysis Lite (SPARFA-Lite), tensor factorization and eXtreme Gradient Boosting (XGBoost). While XGBoost (trained on local machine) outperforms GPT-4 in predictive accuracy, GPT-4-selected XGBoost and its subsequent tuning on the GPT-4 platform demonstrates superior performance compared to local machine execution. Moreover, our investigation into hyper-parameter tuning by GPT-4 versus grid-search suggests comparable performance, albeit with less stability in the automated approach, using XGBoost as the case study. Our study contributes to the field by highlighting the potential of integrating LLMs with traditional machine learning models to enhance predictive accuracy and personalize adult literacy education, setting a foundation for future research in applying LLMs within ITSs.

Liang Zhang, Jionghao Lin, Conrad Borchers et al. · cmu

Learning performance data (e.g., quiz scores and attempts) is significant for understanding learner engagement and knowledge mastery level. However, the learning performance data collected from Intelligent Tutoring Systems (ITSs) often suffers from sparsity, impacting the accuracy of learner modeling and knowledge assessments. To address this, we introduce the 3DG framework (3-Dimensional tensor for Densification and Generation), a novel approach combining tensor factorization with advanced generative models, including Generative Adversarial Network (GAN) and Generative Pre-trained Transformer (GPT), for enhanced data imputation and augmentation. The framework operates by first representing the data as a three-dimensional tensor, capturing dimensions of learners, questions, and attempts. It then densifies the data through tensor factorization and augments it using Generative AI models, tailored to individual learning patterns identified via clustering. Applied to data from an AutoTutor lesson by the Center for the Study of Adult Literacy (CSAL), the 3DG framework effectively generated scalable, personalized simulations of learning performance. Comparative analysis revealed GAN's superior reliability over GPT-4 in this context, underscoring its potential in addressing data sparsity challenges in ITSs and contributing to the advancement of personalized educational technology.

Danielle R. Thomas, Conrad Borchers, Sanjit Kakarla et al. · cmu

The role of multiple-choice questions (MCQs) as effective learning tools has been debated in past research. While MCQs are widely used due to their ease in grading, open response questions are increasingly used for instruction, given advances in large language models (LLMs) for automated grading. This study evaluates MCQs effectiveness relative to open-response questions, both individually and in combination, on learning. These activities are embedded within six tutor lessons on advocacy. Using a posttest-only randomized control design, we compare the performance of 234 tutors (790 lesson completions) across three conditions: MCQ only, open response only, and a combination of both. We find no significant learning differences across conditions at posttest, but tutors in the MCQ condition took significantly less time to complete instruction. These findings suggest that MCQs are as effective, and more efficient, than open response tasks for learning when practice time is limited. To further enhance efficiency, we autograded open responses using GPT-4o and GPT-4-turbo. GPT models demonstrate proficiency for purposes of low-stakes assessment, though further research is needed for broader use. This study contributes a dataset of lesson log data, human annotation rubrics, and LLM prompts to promote transparency and reproducibility.

Conrad Borchers, Yeyu Wang, Shamya Karumbaiah et al.

Learning analytics research increasingly studies classroom learning with AI-based systems through rich contextual data from outside these systems, especially student-teacher interactions. One key challenge in leveraging such data is generating meaningful insights into effective teacher practices. Quantitative ethnography bears the potential to close this gap by combining multimodal data streams into networks of co-occurring behavior that drive insight into favorable learning conditions. The present study uses transmodal ordered network analysis to understand effective teacher practices in relationship to traditional metrics of in-system learning in a mathematics classroom working with AI tutors. Incorporating teacher practices captured by position tracking and human observation codes into modeling significantly improved the inference of how efficiently students improved in the AI tutor beyond a model with tutor log data features only. Comparing teacher practices by student learning rates, we find that students with low learning rates exhibited more hint use after monitoring. However, after an extended visit, students with low learning rates showed learning behavior similar to their high learning rate peers, achieving repeated correct attempts in the tutor. Observation notes suggest conceptual and procedural support differences can help explain visit effectiveness. Taken together, offering early conceptual support to students with low learning rates could make classroom practice with AI tutors more effective. This study advances the scientific understanding of effective teacher practice in classrooms learning with AI tutors and methodologies to make such practices visible.

Danielle R. Thomas, Conrad Borchers, Sanjit Kakarla et al. · cmu

Equity is a core concern of learning analytics. However, applications that teach and assess equity skills, particularly at scale are lacking, often due to barriers in evaluating language. Advances in generative AI via large language models (LLMs) are being used in a wide range of applications, with this present work assessing its use in the equity domain. We evaluate tutor performance within an online lesson on enhancing tutors' skills when responding to students in potentially inequitable situations. We apply a mixed-method approach to analyze the performance of 81 undergraduate remote tutors. We find marginally significant learning gains with increases in tutors' self-reported confidence in their knowledge in responding to middle school students experiencing possible inequities from pretest to posttest. Both GPT-4o and GPT-4-turbo demonstrate proficiency in assessing tutors ability to predict and explain the best approach. Balancing performance, efficiency, and cost, we determine that few-shot learning using GPT-4o is the preferred model. This work makes available a dataset of lesson log data, tutor responses, rubrics for human annotation, and generative AI prompts. Future work involves leveling the difficulty among scenarios and enhancing LLM prompts for large-scale grading and assessment.

Valdemar Švábenský, Conrad Borchers, Elizabeth B. Cloude et al.

In supervised machine learning (SML) research, large training datasets are essential for valid results. However, obtaining primary data in learning analytics (LA) is challenging. Data augmentation can address this by expanding and diversifying data, though its use in LA remains underexplored. This paper systematically compares data augmentation techniques and their impact on prediction performance in a typical LA task: prediction of academic outcomes. Augmentation is demonstrated on four SML models, which we successfully replicated from a previous LAK study based on AUC values. Among 21 augmentation techniques, SMOTE-ENN sampling performed the best, improving the average AUC by 0.01 and approximately halving the training time compared to the baseline models. In addition, we compared 99 combinations of chaining 21 techniques, and found minor, although statistically significant, improvements across models when adding noise to SMOTE-ENN (+0.014). Notably, some augmentation techniques significantly lowered predictive performance or increased performance fluctuation related to random chance. This paper's contribution is twofold. Primarily, our empirical findings show that sampling techniques provide the most statistically reliable performance improvements for LA applications of SML, and are computationally more efficient than deep generation methods with complex hyperparameter settings. Second, the LA community may benefit from validating a recent study through independent replication.

Conrad Borchers, Danielle R. Thomas, Jionghao Lin et al. · cmu

Large Language Models (LLMs) like GPT-4o can help automate text classification tasks at low cost and scale. However, there are major concerns about the validity and reliability of LLM outputs. By contrast, human coding is generally more reliable but expensive to procure at scale. In this study, we propose a hybrid solution to leverage the strengths of both. We combine human-coded data and synthetic LLM-produced data to fine-tune a classical machine learning classifier, distilling both into a smaller BERT model. We evaluate our method on a human-coded test set as a validity measure for LLM output quality. In three experiments, we systematically vary LLM-generated samples' size, variety, and consistency, informed by best practices in LLM tuning. Our findings indicate that augmenting datasets with synthetic samples improves classifier performance, with optimal results achieved at an 80% synthetic to 20% human-coded data ratio. Lower temperature settings of 0.3, corresponding to less variability in LLM generations, produced more stable improvements but also limited model learning from augmented samples. In contrast, higher temperature settings (0.7 and above) introduced greater variability in performance estimates and, at times, lower performance. Hence, LLMs may produce more uniform output that classifiers overfit to earlier or produce more diverse output that runs the risk of deteriorating model performance through information irrelevant to the prediction task. Filtering out inconsistent synthetic samples did not enhance performance. We conclude that integrating human and LLM-generated data to improve text classification models in assessment offers a scalable solution that leverages both the accuracy of human coding and the variety of LLM outputs.

Conrad Borchers, Ryan S. Baker

Algorithmic bias continues to be a key concern of learning analytics. We study the statistical properties of the Absolute Between-ROC Area (ABROCA) metric. This fairness measure quantifies group-level differences in classifier performance through the absolute difference in ROC curves. ABROCA is particularly useful for detecting nuanced performance differences even when overall Area Under the ROC Curve (AUC) values are similar. We sample ABROCA under various conditions, including varying AUC differences and class distributions. We find that ABROCA distributions exhibit high skewness dependent on sample sizes, AUC differences, and class imbalance. When assessing whether a classifier is biased, this skewness inflates ABROCA values by chance, even when data is drawn (by simulation) from populations with equivalent ROC curves. These findings suggest that ABROCA requires careful interpretation given its distributional properties, especially when used to assess the degree of bias and when classes are imbalanced.

Frederik Baucks, Robin Schmucker, Conrad Borchers et al.

Curriculum Analytics (CA) studies curriculum structure and student data to ensure the quality of educational programs. One desirable property of courses within curricula is that they are not unexpectedly more difficult for students of different backgrounds. While prior work points to likely variations in course difficulty across student groups, robust methodologies for capturing such variations are scarce, and existing approaches do not adequately decouple course-specific difficulty from students' general performance levels. The present study introduces Differential Course Functioning (DCF) as an Item Response Theory (IRT)-based CA methodology. DCF controls for student performance levels and examines whether significant differences exist in how distinct student groups succeed in a given course. Leveraging data from over 20,000 students at a large public university, we demonstrate DCF's ability to detect inequities in undergraduate course difficulty across student groups described by grade achievement. We compare major pairs with high co-enrollment and transfer students to their non-transfer peers. For the former, our findings suggest a link between DCF effect sizes and the alignment of course content to student home department motivating interventions targeted towards improving course preparedness. For the latter, results suggest minor variations in course-specific difficulty between transfer and non-transfer students. While this is desirable, it also suggests that interventions targeted toward mitigating grade achievement gaps in transfer students should encompass comprehensive support beyond enhancing preparedness for individual courses. By providing more nuanced and equitable assessments of academic performance and difficulties experienced by diverse student populations, DCF could support policymakers, course articulation officers, and student advisors.

Vincent Aleven, Conrad Borchers, Yun Huang et al.

Intelligent tutoring systems (ITSs) are effective in helping students learn; further research could make them even more effective. Particularly desirable is research into how students learn with these systems, how these systems best support student learning, and what learning sciences principles are key in ITSs. CTAT+Tutorshop provides a full stack integrated platform that facilitates a complete research lifecycle with ITSs, which includes using ITS data to discover learner challenges, to identify opportunities for system improvements, and to conduct experimental studies. The platform includes authoring tools to support and accelerate development of ITS, which provide automatic data logging in a format compatible with DataShop, an independent site that supports the analysis of ed tech log data to study student learnings. Among the many technology platforms that exist to support learning sciences research, CTAT+Tutorshop may be the only one that offers researchers the possibility to author elements of ITSs, or whole ITSs, as part of designing studies. This platform has been used to develop and conduct an estimated 147 research studies which have run in a wide variety of laboratory and real-world educational settings, including K-12 and higher education, and have addressed a wide range of research questions. This paper presents five case studies of research conducted on the CTAT+Tutorshop platform, and summarizes what has been accomplished and what is possible for future researchers. We reflect on the distinctive elements of this platform that have made it so effective in facilitating a wide range of ITS research.

Danielle R. Thomas, Conrad Borchers, Kenneth R. Koedinger

Humans can be notoriously imperfect evaluators. They are often biased, unreliable, and unfit to define "ground truth." Yet, given the surging need to produce large amounts of training data in educational applications using AI, traditional inter-rater reliability (IRR) metrics like Cohen's kappa remain central to validating labeled data. IRR remains a cornerstone of many machine learning pipelines for educational data. Take, for example, the classification of tutors' moves in dialogues or labeling open responses in machine-graded assessments. This position paper argues that overreliance on human IRR as a gatekeeper for annotation quality hampers progress in classifying data in ways that are valid and predictive in relation to improving learning. To address this issue, we highlight five examples of complementary evaluation methods, such as multi-label annotation schemes, expert-based approaches, and close-the-loop validity. We argue that these approaches are in a better position to produce training data and subsequent models that produce improved student learning and more actionable insights than IRR approaches alone. We also emphasize the importance of external validity, for example, by establishing a procedure of validating tutor moves and demonstrating that it works across many categories of tutor actions (e.g., providing hints). We call on the field to rethink annotation quality and ground truth--prioritizing validity and educational impact over consensus alone.