Exploring the Use of Natural Language Processing to Understand Emotions of Trainees and Faculty Regarding Entrustable Professional Activity Assessments
Bibliographic record
Abstract
Background In medical education, artificial intelligence techniques such as natural language processing (NLP) are starting to be used to capture and analyze emotions through written text. Objective To explore the application of NLP techniques to understand resident and faculty emotions related to entrustable professional activity (EPA) assessments. Methods Open-ended text data from a survey on emotions toward EPA assessments were analyzed. Respondents were residents and faculty from pediatrics (Peds), general surgery (GS), and emergency medicine (EM), recruited for a larger emotions study in 2023. Participants wrote about their emotions related to receiving/completing EPA assessments. We analyzed the frequency of words rated as positive via a validated sentiment lexicon used in NLP studies. Specifically, we were interested if the count of positive words varied as a function of group membership (faculty, resident), specialty (Peds, GS, EM), gender (man, woman, nonbinary), or visible minority status (yes, no, omit). Results A total of 66 text responses (30 faculty, 36 residents) contained text data useful for sentiment analysis. We analyzed the difference in the count of words categorized as positive across group, specialty, gender, and being a visible minority. Specialty was the only category revealing significant differences via a bootstrapped Poisson regression model with GS responses containing fewer positive words than EM responses. Conclusions By analyzing text data to understand emotions of residents and faculty through an NLP approach, we identified differences in EPA assessment-related emotions of residents versus faculty, and differences across specialties.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.029 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 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 source (direct Gemma or distilled Codex), 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".