Automatic Detection of Usability Problem Encounters in Think-aloud Sessions
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Think-aloud protocols are a highly valued usability testing method for identifying usability problems. Despite the value of conducting think-aloud usability test sessions, analyzing think-aloud sessions is often time-consuming and labor-intensive. Consequently, previous research has urged the community to develop techniques to support fast-paced analysis. In this work, we took the first step to design and evaluate machine learning (ML) models to automatically detect usability problem encounters based on users’ verbalization and speech features in think-aloud sessions. Inspired by recent research that shows subtle patterns in users’ verbalizations and speech features tend to occur when they encounter problems, we examined whether these patterns can be utilized to improve the automatic detection of usability problems. We first conducted and recorded think-aloud sessions and then examined the effect of different input features, ML models, test products, and users on usability problem encounters detection. Our work uncovers several technical and user interface design challenges and sets a baseline for automating usability problem detection and integrating such automation into UX practitioners’ workflow.
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.
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 | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| 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 it