Andrew Katz

Andrew Katz, Gabriella Coloyan Fleming, Joyce Main

This paper aims to answer one central question: to what extent can open-source generative text models be used in a workflow to approximate thematic analysis in social science research? To answer this question, we present the Generative AI-enabled Theme Organization and Structuring (GATOS) workflow, which uses open-source machine learning techniques, natural language processing tools, and generative text models to facilitate thematic analysis. To establish validity of the method, we present three case studies applying the GATOS workflow, leveraging these models and techniques to inductively create codebooks similar to traditional procedures using thematic analysis. Specifically, we investigate the extent to which a workflow comprising open-source models and tools can inductively produce codebooks that approach the known space of themes and sub-themes. To address the challenge of gleaning insights from these texts, we combine open-source generative text models, retrieval-augmented generation, and prompt engineering to identify codes and themes in large volumes of text, i.e., generate a qualitative codebook. The process mimics an inductive coding process that researchers might use in traditional thematic analysis by reading text one unit of analysis at a time, considering existing codes already in the codebook, and then deciding whether or not to generate a new code based on whether the extant codebook provides adequate thematic coverage. We demonstrate this workflow using three synthetic datasets from hypothetical organizational research settings: a study of teammate feedback in teamwork settings, a study of organizational cultures of ethical behavior, and a study of employee perspectives about returning to their offices after the pandemic. We show that the GATOS workflow is able to identify themes in the text that were used to generate the original synthetic datasets.

Xiomara Gonzalez, Gabriella Coloyan Fleming, Andrew Katz et al.

Written reflection assignments give students valuable opportunities for critical self-assessment, meaning making, and learning processing. Additionally, such reflections provide rich data for qualitative education research. However, qualitative data can be time-consuming to analyze. It is even more time-intensive to qualitatively compare findings between different groups of participants, usually limiting comparison to, at most, one variable (e.g., binary gender). Large language models (LLMs) have recently begun to be critically evaluated for use as qualitative research assistants. Using a longitudinal case of written student reflections (n=151) from a study abroad program, we investigate how LLM-assisted sentiment analysis can enable longitudinal mixed-methods research combining computational and thematic analyses. First, statistical testing is used to quantitatively compare sentiment differences according to seven different student identity/lived experience variables. Then, these results inform qualitative data analysis to investigate the reasons underlying these differences. For the case of undergraduate students studying abroad, we found that prior experience living abroad was the only personal variable impacting students' sentiments of their verbal language and communication behaviors. This workflow has implications for how qualitative researchers can more easily probe multiple variables when comparing participants from different demographic groups.

Nolan Platt, Sehrish Nizamani, Alp Tural et al.

Understanding student engagement usually requires time-consuming manual observation or invasive recording that raises privacy concerns. We present a privacy-preserving pipeline that analyzes classroom videos to extract insights about student attention, without storing any identifiable footage. Our system runs on a single GPU, using OpenPose for skeletal extraction and Gaze-LLE for visual attention estimation. Original video frames are deleted immediately after pose extraction, thus only geometric coordinates (stored as JSON) are retained, ensuring compliance with FERPA. The extracted pose and gaze data is processed by QwQ-32B-Reasoning, which performs zero-shot analysis of student behavior across lecture segments. Instructors access results through a web dashboard featuring attention heatmaps and behavioral summaries. Our preliminary findings suggest that LLMs may show promise for multimodal behavior understanding, although they still struggle with spatial reasoning about classroom layouts. We discuss these limitations and outline directions for improving LLM spatial comprehension in educational analytics contexts.

Andrew Katz, Mitchell Gerhardt, Michelle Soledad

Feedback is a critical aspect of improvement. Unfortunately, when there is a lot of feedback from multiple sources, it can be difficult to distill the information into actionable insights. Consider student evaluations of teaching (SETs), which are important sources of feedback for educators. They can give instructors insights into what worked during a semester. A collection of SETs can also be useful to administrators as signals for courses or entire programs. However, on a large scale as in high-enrollment courses or administrative records over several years, the volume of SETs can render them difficult to analyze. In this paper, we discuss a novel method for analyzing SETs using natural language processing (NLP) and large language models (LLMs). We demonstrate the method by applying it to a corpus of 5,000 SETs from a large public university. We show that the method can be used to extract, embed, cluster, and summarize the SETs to identify the themes they express. More generally, this work illustrates how to use the combination of NLP techniques and LLMs to generate a codebook for SETs. We conclude by discussing the implications of this method for analyzing SETs and other types of student writing in teaching and research settings.

Andrew Katz

The minimal pairs paradigm of comparing model probabilities for contrasting completions has proven useful for evaluating linguistic knowledge in language models, yet its application has largely been confined to binary grammaticality judgments over syntactic phenomena. Additionally, standard prompting-based evaluation requires expensive text generation, may elicit post-hoc rationalizations rather than model judgments, and discards information about model uncertainty. We address both limitations by extending surprisal-based evaluation from binary grammaticality contrasts to ordinal-scaled classification and scoring tasks across multiple domains. Rather than asking models to generate answers, we measure the information-theoretic "surprise" (negative log probability) they assign to each position on rating scales (e.g., 1-5 or 1-9), yielding full surprisal curves that reveal both the model's preferred response and its uncertainty via entropy. We explore this framework across four domains: social-ecological-technological systems classification, causal statement identification (binary and scaled), figurative language detection, and deductive qualitative coding. Across these domains, surprisal curves produce interpretable classification signals with clear minima near expected ordinal scale positions, and entropy over the completion tended to distinguish genuinely ambiguous items from easier items.

Andrew Katz, Siqing Wei, Gaurav Nanda et al.

Teamwork is a critical component of many academic and professional settings. In those contexts, feedback between team members is an important element to facilitate successful and sustainable teamwork. However, in the classroom, as the number of teams and team members and frequency of evaluation increase, the volume of comments can become overwhelming for an instructor to read and track, making it difficult to identify patterns and areas for student improvement. To address this challenge, we explored the use of generative AI models, specifically ChatGPT, to analyze student comments in team based learning contexts. Our study aimed to evaluate ChatGPT's ability to accurately identify topics in student comments based on an existing framework consisting of positive and negative comments. Our results suggest that ChatGPT can achieve over 90\% accuracy in labeling student comments, providing a potentially valuable tool for analyzing feedback in team projects. This study contributes to the growing body of research on the use of AI models in educational contexts and highlights the potential of ChatGPT for facilitating analysis of student comments.

Chenhan Yuan, Ryan Rossi, Andrew Katz et al.

The models of n-ary cross sentence relation extraction based on distant supervision assume that consecutive sentences mentioning n entities describe the relation of these n entities. However, on one hand, this assumption introduces noisy labeled data and harms the models' performance. On the other hand, some non-consecutive sentences also describe one relation and these sentences cannot be labeled under this assumption. In this paper, we relax this strong assumption by a weaker distant supervision assumption to address the second issue and propose a novel sentence distribution estimator model to address the first problem. This estimator selects correctly labeled sentences to alleviate the effect of noisy data is a two-level agent reinforcement learning model. In addition, a novel universal relation extractor with a hybrid approach of attention mechanism and PCNN is proposed such that it can be deployed in any tasks, including consecutive and nonconsecutive sentences. Experiments demonstrate that the proposed model can reduce the impact of noisy data and achieve better performance on general n-ary cross sentence relation extraction task compared to baseline models.

Chenhan Yuan, Ryan Rossi, Andrew Katz et al.

This paper focuses on the problem of unsupervised relation extraction. Existing probabilistic generative model-based relation extraction methods work by extracting sentence features and using these features as inputs to train a generative model. This model is then used to cluster similar relations. However, these methods do not consider correlations between sentences with the same entity pair during training, which can negatively impact model performance. To address this issue, we propose a Clustering-based Unsupervised generative Relation Extraction (CURE) framework that leverages an "Encoder-Decoder" architecture to perform self-supervised learning so the encoder can extract relation information. Given multiple sentences with the same entity pair as inputs, self-supervised learning is deployed by predicting the shortest path between entity pairs on the dependency graph of one of the sentences. After that, we extract the relation information using the well-trained encoder. Then, entity pairs that share the same relation are clustered based on their corresponding relation information. Each cluster is labeled with a few words based on the words in the shortest paths corresponding to the entity pairs in each cluster. These cluster labels also describe the meaning of these relation clusters. We compare the triplets extracted by our proposed framework (CURE) and baseline methods with a ground-truth Knowledge Base. Experimental results show that our model performs better than state-of-the-art models on both New York Times (NYT) and United Nations Parallel Corpus (UNPC) standard datasets.