Using LLMs to Analyze Antecedent, Behavior and Consequence Narrative Recordings in Behavioral Health Science
Bibliographic record
Abstract
The Antecedent, Behavior, and Consequence narrative recordings is a recognized approach for data collection and analysis in conducting a Functional Behavior Assessment. Currently, this method relies on a manual process that necessitates detailed descriptive analyses of the behavior in question, which must be performed by certified professionals, such as Board Certified Behavior Analysts (BCBAs). Only 4.7% of Board Certified Behavior Analysts (BCBAs) reside outside US and Canada making it difficult to conduct a Functional Behavior Assessment outside these two countries. Within the US, systemic challenges, including insufficient funding within public education and operational inefficiencies, have contributed to significant delays in executing Functional Behavior Assessments with wait times often extending for several months. This paper proposes the utilization of Chain of Thought prompting with Large Language Models to analyze data derived from Antecedent, Behavior, and Consequence narrative recordings. Through our proposed approach, we achieve an 78% accuracy in identifying antecedent labels and 70% accuracy for consequence labels from narrative recordings, demonstrating that this method produces scalable and accurate results. To ensure the generalizability of our approach, especially in scenarios with limited data, we employed data augmentation techniques and Leave One Out Cross Validation (LOOCV). Our model significantly outperformed traditional machine learning algorithms like Support Vector Machines and Random Forest, demonstrating its robustness and potential for real-world applications. By reducing operational inefficiencies, this approach can significantly improve the accessibility of behavioral healthcare for traditionally marginalized populations such as low-income, rural, minority, and other disadvantaged communities across the world. Furthermore, we aim to position this paper as a catalyst for advancing the use of Large Language Models and Natural Language Processing in the field of behavioral health, paving the way for further research and development.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.004 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".