Generating synthetic data for CALL research with GenAI: A proof-of-concept study
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
Popular tools like ChatGPT have placed generative artificial intelligence (GenAI) in the spotlight in recent years. One use of GenAI tools is to generate simulated data—or synthetic data—when the full scope of the required microdata is unavailable. Despite suggestions for educational researchers to use synthetic data, little (if any) computer-assisted language learning (CALL) research has used synthetic data thus far. This study addresses this research gap by exploring the possibility of using synthetic datasets in CALL. The publicly available dataset resembles a typical study with a small sample size ( n = 55) performed using a CALL platform. Two synthetic datasets are generated from the original datasets using the synthpop package and generative adversarial networks (GAN) in R (via the RGAN package), which are both common synthetic data generation methods. This study evaluates the synthetic datasets by (a) comparing the distribution between the synthetic and original datasets, (b) examining the model parameters of the rebuilt linear models using the synthetic and original datasets, and (c) examining the privacy disclosure metrics. The results suggest that synthpop better represents the original data and preserves privacy. Notably, the GAN-generated dataset does not produce satisfactory results. This demonstrates GAN’s key challenges alongside the potential benefits of generating synthetic data with synthpop .
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 | 0.038 | 0.032 |
| 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 | 0.000 | 0.000 |
| Open science | 0.003 | 0.002 |
| 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