Automated scoring in the era of artificial intelligence: An empirical study with Turkish essays
Bibliographic record
Abstract
Automated scoring (AS) has gained significant attention as a tool to enhance the efficiency and reliability of assessment processes. Yet, its application in under-represented languages, such as Turkish, remains limited. This study addresses this gap by empirically evaluating AS for Turkish using a zero-shot approach with a rubric powered by OpenAI's GPT-4o. A dataset of 590 essays written by learners of Turkish as a second language was scored by professional human raters and an artificial intelligence (AI) model integrated via a custom-built interface. The scoring rubric, grounded in the Common European Framework of Reference for Languages, assessed six dimensions of writing quality. Results revealed a strong alignment between human and AI scores with a Quadratic Weighted Kappa of 0.72, Pearson correlation of 0.73, and an overlap measure of 83.5 %. Analysis of rater effects showed minimal influence on score discrepancies, though factors such as experience and gender exhibited modest effects. These findings demonstrate the potential of AI-driven scoring in Turkish, offering valuable insights for broader implementation in under-represented languages, such as the possible source of disagreements between human and AI scores. Conclusions from a specific writing task with a single human rater underscore the need for future research to explore diverse inputs and multiple raters.
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 | 0.015 | 0.110 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.003 | 0.004 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.003 | 0.002 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 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".