Summarizing Students’ Free Responses for an Introductory Algebra-Based Physics Course Survey Using Cluster and Sentiment Analysis
Bibliographic record
Abstract
In Physics Higher Education (PHE), Student Evaluation of Teaching (SET) surveys are widely used to collect students’ feedback on courses and instructions. In our research, we propose a more efficient way to summarize students’ free responses from the Student Assessment of their Learning Gains (SALG) survey [1], a form of the SET survey, of an algebra-based introductory physics course at a large Canadian research university. Specifically, we use cluster and sentiment analysis methods such as K-means [2] and Valence Aware Dictionary for sEntiment Reasoning (VADER) [3] to summarize students’ free responses. For cluster analysis, we extract popular keywords and summaries of responses in different clusters that reflect students’ dominant opinions toward each aspect of the course. Notably, we obtain an average silhouette coefficient of 0.480. In addition, we analyze sentiments in students’ free responses that are determined through applying VADER. Intriguingly, we see that VADER (micro F1 = 0.57, macro F1 = 0.55) can better classify responses with positive (F1 = 0.62) and neutral sentiment (F1 = 0.59). However, evident disagreements arise with negative sentiment responses (F1 = 0.42). In addition, our research suggests that some Likert-scale summaries deviate from the sentiment of free response summaries due to the limitations of Likert-scale responses. By creating various visualizations, we discover that Natural Language Processing (NLP) methods, such as cluster and sentiment analysis, effectively summarize students’ free responses, with several limitations.
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 | 0.018 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.004 | 0.003 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
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".