Pith. sign in

REVIEW 3 major objections 5 minor 38 references

Does the Prompt-based Large Language Model Recognize Students' Demographics and Introduce Bias in Essay Scoring?

T0 review · 3 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

Pith's one-line read GPT-4o can infer first-language background from essays and scores non-native writers less fairly when it does

desk verdict First real test of the demographic-inference/scoring-bias link in prompt-based AES, with a solid pooled interaction for first-language background, but the central causal claim is not identified because 'Correct' is a proxy for essay surface features. read the letter →

arxiv 2504.21330 v1 pith:FMAU6LTN submitted 2025-04-30 cs.CL

classification cs.CL
keywords automatedessayscoringprompt-basedlargelanguagemodeldemographicinferencefirst-languagebackgroundalgorithmicfairnessGPT-4oeducationalNLPbias
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper asks whether a prompt-based large language model, GPT-4o, can read students' demographics from their essays and whether that recognition changes the fairness of automated scoring. On six independent writing topics from an argumentative-essay corpus, the model inferred first-language background with accuracy around $0.75$–$0.87$ when it gave a definite answer, and its scoring error for non-native English speakers increased when it correctly identified them as non-native. The stakes are concrete because prompt-based essay scoring is accessible to educators with no machine-learning expertise, so a hidden demographic penalty could enter real classrooms unnoticed. The paper also finds little gender bias in this setup, even though the model's gender guesses were accurate on the small share of essays where it committed to a prediction.

What carries the argument

The central object is the pair of independent GPT-4o prompts: a demographic-inference prompt (gender or first-language background, each with an 'Uncertain' option) and a rubric-based scoring prompt with few-shot examples and chain-of-thought steps. The argument is carried by splitting demographic predictions into 'Correct' versus 'Unreliable' (wrong or uncertain) and then fitting a weighted multivariate regression of scoring error on demographics, correctness, and the correctness-by-language interaction, with inverse probability weighting to keep the minority group represented. The positive, significant interaction coefficient is the load-bearing quantity: it is what connects successful demographic recognition to larger scoring errors for non-native writers.

What would settle it

Rerun the weighted regression with an added control for surface grammatical error rate, or score human-corrected transcripts of the non-native essays; if the correctness-by-language coefficient drops to zero, the apparent demographic bias is actually error-pattern bias and the paper's conclusion fails.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that GPT-4o's scoring fairness depends on whether the model can recognize the writer's first-language background. For essays it classified as native or non-native, the model reached near-total coverage and roughly $0.75$–$0.87$ accuracy, and the weighted regression of absolute scoring error on demographics, prediction correctness, and their interaction produced a positive and significant correctness-by-language coefficient of $0.502$ (SE $0.074$) overall. That coefficient means non-native English speakers are scored with bigger errors exactly when the model has correctly labeled them as non-native. Fairness metrics (OSA, OSD, CSD, MAED) show more pronounced and more often statistically significant bias in the correctly-predicted group than in the unreliable group. The same relationship does not appear for gender, which the paper attributes to gender information being less directly accessible through prompts and to the model's low commitment rate on gender judgments.

Load-bearing premise

The load-bearing premise is that the 'Correct' versus 'Unreliable' split cleanly measures the model's demographic recognition; if being 'Correct' actually tracks essay traits such as grammatical error patterns or topic, the positive correctness-by-language interaction in Section 4.2 could reflect those traits rather than demographic bias, and the central claim would not follow.

Editorial extensions

If this is right

  • If the result holds, prompt-based essay scoring with GPT-4o will systematically disadvantage non-native English writers whenever the essay text reveals their language background.
  • Because first-language coverage is near 100 percent, this penalty applies to almost all non-native writers in the analyzed essays, not to a small subset.
  • Debiasing strategies developed for fine-tuned scoring models do not transfer directly to prompt-only tools, so fair use requires new mitigation such as deliberately diverse few-shot examples.
  • The absence of consistent gender bias shows that demographic fairness must be checked attribute by attribute; a model can be fair on gender while biased on language background, or vice versa.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A direct causal reading of the interaction implies that stripping away surface error patterns, for example by editing non-native essays into standard English before scoring, should reduce the bias; the paper does not run that test.
  • The gender null result may be an artifact of low coverage: with only 4–13 percent of essays receiving a definite gender prediction, the analysis excludes most of the data, so a forced-choice prompt could reveal gender bias that this design misses.
  • The demographic-inference prompt itself could serve as an audit tool, letting a school or platform measure how accurately any LLM guesses student attributes and use the correctness-by-language coefficient as a pre-deployment fairness check.
  • If the bias mechanism is demographic recognition rather than essay quality, then telling the model the author is non-native, or removing such cues, should shift the interaction term and give a cheap experimental handle for future studies.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. This manuscript investigates whether a prompt-based large language model (GPT-4o) can infer students' gender and first-language background from their essays, and whether the accuracy of those demographic inferences is associated with bias in automated essay scoring. Using six essay sets from the PERSUADE 2.0 corpus, the authors prompt GPT-4o to predict demographics and to score essays, then compare fairness metrics and fit inverse-probability-weighted regressions with demographic group, correctness of demographic prediction, and their interaction. The main reported findings are that first-language background can be inferred with roughly 79–87% accuracy on covered cases, and that the regression interaction Correctness×Language is positive overall (coefficient 0.502, SE 0.074), interpreted as larger scoring error for non-native English speakers when the model correctly identifies them as non-native. For gender, coverage is very low (4–13%) and no consistent relationship is found.

Significance. If the central claim held, the study would provide an actionable connection between demographic inferability and scoring bias in prompt-based AES, going beyond prior work on fine-tuned models and suggesting concrete debiasing levers for widely used LLM APIs. The paper has notable strengths: a large public corpus, a transparent prompt design, robustness checks via five paraphrased prompts, and clearly specified regression models. The demographic inference results for RQ1 are a solid empirical contribution, especially the finding that first-language background is more detectable than gender. However, the load-bearing claim (iii) is an interpretation of a non-experimental comparison between 'Correct' and 'Unreliable' groups, and the manuscript does not yet rule out essay-level linguistic confounds. The significance of the paper is therefore conditional on either additional covariate-adjusted analysis or a more modest, descriptive framing of the main finding.

major comments (3)
  1. [§3.3 and Table 4] The Correct/Unreliable dichotomy does not isolate demographic recognition. As defined, 'Unreliable' merges incorrect predictions and 'Uncertain'; for first-language background the coverage is 97–99% (Table 2), so 'Unreliable' is almost entirely essays the model misclassified. A non-native essay is misclassified as native precisely when its linguistic surface resembles native writing, and that same surface property is likely associated with how the scoring model errs, independent of any demographic 'awareness'. The regression in Table 4 adjusts for Language, Correctness, and their interaction but includes no essay-level covariates (e.g., lexical sophistication, grammatical error density, topic) in the overall model. The positive Correctness×Language coefficient may therefore reflect selection on surface properties rather than a demographic-recognition mechanism. Please add covariates or explicitly report the analysis as a descriptive association, and revise the causal phrasing in the abstract and Section 5 accordingly.
  2. [§4.2, Table 4] The interaction is not consistent across essay sets. Essay Set 2 shows a significant negative interaction (-0.490, SE 0.203), Sets 1 and 4 are non-significant, and the positive overall estimate appears driven largely by Set 6, the largest set. The abstract's unconditional claim (iii) is not supported by the per-set results. A heterogeneity test, a mixed-effects model with essay-set random slopes, or a meta-analytic summary would clarify whether a general phenomenon exists or whether the effect is topic-specific. Without such an analysis, the overall interaction should be interpreted with caution.
  3. [§4.2, Tables 3 and 4] The paper does not correct for multiple testing across six essay sets, two demographic attributes, four fairness metrics, and a large family of regression coefficients. For example, the gender interaction terms in Sets 3 and 5, and some of the OSA/CSD entries in Table 3, could be chance findings. Please report adjusted p-values (e.g., Benjamini-Hochberg) or at minimum disclose the total number of tests and argue why the pattern of significance is robust. This is especially important because the main claim rests on a single interaction coefficient in the overall model.
minor comments (5)
  1. [§3.5] Please clarify how majority voting was applied for demographic predictions when the five paraphrased prompts produced a tie, and whether the same repeated prompts were used for scoring (which was averaged). Reporting agreement among the five runs would also strengthen the robustness claim.
  2. [§3.3] The inverse probability weighting is not fully specified. State how the weights were computed (e.g., inverse group frequency within each essay set) and indicate whether weighting was applied to the QWK and MAED calculations or only to the regressions and fairness regression-based metrics.
  3. [§4.1] The text says the model achieved 'an accuracy of about 0.79–0.86' for first-language background, but Table 2 shows accuracy values from 0.753 to 0.868. The reported range is inconsistent with the table; please correct the text to match the data.
  4. [§3.5] The sentence 'Each experiment was run five times with paraphrased prompts by other prompt-based LLMs (i.e., Gemini and Claude)' is ambiguous. It is unclear whether the paraphrases were generated by Gemini/Claude or whether the experiments themselves were executed with those models. Please rephrase.
  5. [Table 3] The table would benefit from a note explaining the direction of MAED and OSD (which group is over- or under-scored) and why a negative value indicates larger error for a particular group. As written, the interpretation of negative versus positive values across rows is not transparent to a reader unfamiliar with the metrics.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the demographic-inference and scoring-error analyses are estimated from external data rather than derived from fitted inputs.

full rationale

The paper's derivation chain is empirical and self-contained. The demographic prediction accuracy (Section 4.1, Table 2) is evaluated against ground-truth labels from the PERSUADE 2.0 corpus, and the scoring-error regression (Section 4.2, Table 4) uses human rater scores as the dependent variable, with the LLM's demographic-prediction correctness (Correct/Unreliable) and student demographics as predictors. The key interaction term (Correctness*Language, coefficient 0.502, SE 0.074, Overall) is estimated from the data, not imposed by construction; "Correct" is defined by agreement with ground-truth in a separate prompt session, not by the scoring outcome. The only self-citations ([25] and [34]) supply context and a fairness-evaluation framework, but the central claims do not reduce to those citations: the results are reported against external benchmarks (human scores, demographic labels) and would stand or fall on the data regardless of the cited prior work. No fitted parameter is renamed as a prediction, and no equation equates the input with the output. Therefore the paper exhibits no significant circularity.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The study is observational and uses an existing model and dataset. No hand-fitted constants or new theoretical entities are introduced. The main burden is on the interpretive assumption that prediction correctness is not confounded with essay proficiency.

assumptions (3)
  • domain assumption Ground-truth demographics and human rater scores in PERSUADE 2.0 are accurate.
    All measurement of prediction accuracy and scoring error treats these labels as ground truth; if labels are noisy, reported accuracies and bias estimates are biased. Invoked in Section 3.1.
  • ad hoc to paper The Correct versus Unreliable grouping isolates the model's demographic recognition rather than essay quality or topic.
    The paper defines 'Correct' as any definite prediction equal to ground truth and bundles wrong predictions with 'Uncertain'; the regression then interprets interaction terms as if correctness is an independent signal. This is the central interpretive assumption in Section 3.3 and Section 4.2.
  • domain assumption The six retained essay sets represent the broader AES setting.
    Topics lacking demographic info or source articles were dropped; results may not generalize to all PERSUADE topics or other datasets. Stated in Section 3.1.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Does the Prompt-based Large Language Model Recognize Students' Demographics and Introduce Bias in Essay Scoring?." pith.science (2026). https://pith.science/paper/FMAU6LTN

@misc{pith2026250421330,
  author       = {Pith},
  title        = {Pith review of: Does the Prompt-based Large Language Model Recognize Students' Demographics and Introduce Bias in Essay Scoring?},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FMAU6LTN}},
  note         = {Machine review of arXiv:2504.21330}
}
read the original abstract

Large Language Models (LLMs) are widely used in Automated Essay Scoring (AES) due to their ability to capture semantic meaning. Traditional fine-tuning approaches required technical expertise, limiting accessibility for educators with limited technical backgrounds. However, prompt-based tools like ChatGPT have made AES more accessible, enabling educators to obtain machine-generated scores using natural-language prompts (i.e., the prompt-based paradigm). Despite advancements, prior studies have shown bias in fine-tuned LLMs, particularly against disadvantaged groups. It remains unclear whether such biases persist or are amplified in the prompt-based paradigm with cutting-edge tools. Since such biases are believed to stem from the demographic information embedded in pre-trained models (i.e., the ability of LLMs' text embeddings to predict demographic attributes), this study explores the relationship between the model's predictive power of students' demographic attributes based on their written works and its predictive bias in the scoring task in the prompt-based paradigm. Using a publicly available dataset of over 25,000 students' argumentative essays, we designed prompts to elicit demographic inferences (i.e., gender, first-language background) from GPT-4o and assessed fairness in automated scoring. Then we conducted multivariate regression analysis to explore the impact of the model's ability to predict demographics on its scoring outcomes. Our findings revealed that (i) prompt-based LLMs can somewhat infer students' demographics, particularly their first-language backgrounds, from their essays; (ii) scoring biases are more pronounced when the LLM correctly predicts students' first-language background than when it does not; and (iii) scoring error for non-native English speakers increases when the LLM correctly identifies them as non-native.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 31 canonical work pages

  1. [1]

    Creative Edu cation 15(7), 1499–1523 (2024)

    Binhammad, M.H.Y., Othman, A., Abuljadayel, L., Al Mheir i, H., Alkaabi, M., Almarri, M.: Investigating how generative ai can create per sonalized learning ma- terials tailored to individual student needs. Creative Edu cation 15(7), 1499–1523 (2024)

  2. [2]

    Ai magazine 25(3), 27–27 (2004)

    Burstein, J., Chodorow, M., Leacock, C.: Automated essay evaluation: The crite- rion online writing service. Ai magazine 25(3), 27–27 (2004)

  3. [3]

    arXiv preprint arXiv:2310.14735 (2023)

    Chen, B., Zhang, Z., Langrené, N., Zhu, S.: Unleashing the potential of prompt engineering in large language models: a comprehensive revi ew. arXiv preprint arXiv:2310.14735 (2023)

  4. [4]

    arXiv preprint arXiv:2410.11123 (2024)

    Chen, E., Wang, D., Xu, L., Cao, C., Fang, X., Lin, J.: A syst ematic review on prompt engineering in large language models for k-12 stem education. arXiv preprint arXiv:2410.11123 (2024)

  5. [5]

    In: Proceedings of the 2013 conference on empirical me thods in natural language processing

    Chen, H., He, B.: Automated essay scoring by maximizing hu man-machine agree- ment. In: Proceedings of the 2013 conference on empirical me thods in natural language processing. pp. 1741–1752 (2013)

  6. [6]

    Assessing Writing 61, 100865 (2024)

    Crossley, S.A., Tian, Y., Baffour, P., Franklin, A., Benne r, M., Boser, U.: A large- scale corpus for assessing written argumentation: Persuad e 2.0. Assessing Writing 61, 100865 (2024)

  7. [7]

    research notes

    Deane, P.: The importance of assessing student writing an d improving writing instruction. research notes. Educational Testing Service (2022)

  8. [8]

    I n: 16th International Conference on Educational Data Mining, EDM 2023

    Doewes, A., Kurdhi, N., Saxena, A.: Evaluating quadratic weighted kappa as the standard performance metric for automated essay scoring. I n: 16th International Conference on Educational Data Mining, EDM 2023. pp. 103–11 3. International Educational Data Mining Society (IEDMS) (2023)

Show all 38 references
  1. [9]

    In: Proceedings of the 21s t conference on com- putational natural language learning (CoNLL 2017)

    Dong, F., Zhang, Y., Yang, J.: Attention-based recurrent convolutional neural net- work for automatic essay scoring. In: Proceedings of the 21s t conference on com- putational natural language learning (CoNLL 2017). pp. 153 –162 (2017)

  2. [10]

    , Kadavath, S., Mann, B., Perez, E., Schiefer, N., Ndousse, K., et al.: Red teaming language models to reduce harms: Methods, scaling behaviors, and lessons lear ned

    Ganguli, D., Lovitt, L., Kernion, J., Askell, A., Bai, Y. , Kadavath, S., Mann, B., Perez, E., Schiefer, N., Ndousse, K., et al.: Red teaming language models to reduce harms: Methods, scaling behaviors, and lessons lear ned. arXiv preprint arXiv:2209.07858 (2022)

  3. [11]

    Science & Edu cation pp

    García-Méndez, S., de Arriba-Pérez, F., Somoza-López, M.d.C.: A review on the use of large language models as virtual tutors. Science & Edu cation pp. 1–16 (2024)

  4. [12]

    Canadian Journal o f Education/Revue cana- dienne de l’éducation pp

    Jones, S., Myhill, D.: Discourses of difference? examini ng gender differences in linguistic characteristics of writing. Canadian Journal o f Education/Revue cana- dienne de l’éducation pp. 456–482 (2007)

  5. [13]

    In: Proceedings of t he 19th Workshop on Innovative Use of NLP for Building Educational Application s (BEA 2024)

    Kwako, A., Ormerod, C.: Can language models guess your id entity? analyzing demographic biases in ai essay scoring. In: Proceedings of t he 19th Workshop on Innovative Use of NLP for Building Educational Application s (BEA 2024). pp. 78–86 (2024)

  6. [14]

    In: 2021 International Conference on Computer, Information an d Telecommunication Systems (CITS)

    Lagakis, P., Demetriadis, S.: Automated essay scoring: A review of the field. In: 2021 International Conference on Computer, Information an d Telecommunication Systems (CITS). pp. 1–6. IEEE (2021)

  7. [15]

    In: Pro ceedings of the 2022 CHI conference on human factors in computing systems

    Lee, M., Liang, P., Yang, Q.: Coauthor: Designing a human -ai collaborative writing dataset for exploring language model capabilities. In: Pro ceedings of the 2022 CHI conference on human factors in computing systems. pp. 1–19 ( 2022)

  8. [16]

    Comp uters and Education: Artificial Intelligence 4, 100140 (2023) 14 K

    Li, Y., Sha, L., Yan, L., Lin, J., Raković, M., Galbraith, K., Lyons, K., Gašević, D., Chen, G.: Can large language models write reflectively. Comp uters and Education: Artificial Intelligence 4, 100140 (2023) 14 K. Yang et al

  9. [17]

    In: Inter- national Conference on Artificial Intelligence in Educatio n

    Litman, D., Zhang, H., Correnti, R., Matsumura, L.C., Wa ng, E.: A fairness eval- uation of automated methods for scoring text evidence usage in writing. In: Inter- national Conference on Artificial Intelligence in Educatio n. pp. 255–267. Springer (2021)

  10. [18]

    In: Proceedings of the fo urteenth workshop on innovative use of NLP for building educational application s

    Loukina, A., Madnani, N., Zechner, K.: The many dimensio ns of algorithmic fair- ness in educational applications. In: Proceedings of the fo urteenth workshop on innovative use of NLP for building educational application s. pp. 1–10 (2019)

  11. [19]

    Mayfield, E., Black, A.W.: Should you fine-tune bert for au tomated essay scoring? In: Proceedings of the Fifteenth Workshop on Innovative Use of NLP for Building Educational Applications. pp. 151–162 (2020)

  12. [20]

    ACM computing survey s (CSUR) 54(6), 1–35 (2021)

    Mehrabi, N., Morstatter, F., Saxena, N., Lerman, K., Gal styan, A.: A survey on bias and fairness in machine learning. ACM computing survey s (CSUR) 54(6), 1–35 (2021)

  13. [21]

    , Lieb, M., White, J.: Evaluating persona prompting for question answering tasks

    Olea, C., Tucker, H., Phelan, J., Pattison, C., Zhang, S. , Lieb, M., White, J.: Evaluating persona prompting for question answering tasks . In: Proceedings of th e 10th international conference on artificial intelligen ce and soft computing, Sydney, Australia (2024)

  14. [22]

    arXiv preprint arXiv:1909.09482 (2019)

    Rodriguez, P.U., Jafari, A., Ormerod, C.M.: Language mo dels and automated essay scoring. arXiv preprint arXiv:1909.09482 (2019)

  15. [23]

    arXiv preprint arXiv:2402.07927 (2024)

    Sahoo, P., Singh, A.K., Saha, S., Jain, V., Mondal, S., Ch adha, A.: A systematic survey of prompt engineering in large language models: Tech niques and applica- tions. arXiv preprint arXiv:2402.07927 (2024)

  16. [24]

    In: Proceedings of the 19th Workshop o n Innovative Use of NLP for Building Educational Applications (BEA 2024)

    Schaller, N.J., Ding, Y., Horbach, A., Meyer, J., Jansen , T.: Fairness in automated essay scoring: A comparative analysis of algorithms on germ an learner essays from secondary education. In: Proceedings of the 19th Workshop o n Innovative Use of NLP for Building Educational ...

  17. [25]

    In : International Conference on Computational Linguistics 2022

    Sha, L., Li, Y., Gasevic, D., Chen, G.: Bigger data or fair er data?: augmenting bert via active sampling for educational text classification. In : International Conference on Computational Linguistics 2022. pp. 1275–1285. Associa tion for Computational Linguistics (ACL) (2022)

  18. [26]

    arXiv preprint arXiv:2404.15845 (2024)

    Stahl, M., Biermann, L., Nehring, A., Wachsmuth, H.: Exp loring llm prompt- ing strategies for joint essay scoring and feedback generat ion. arXiv preprint arXiv:2404.15845 (2024)

  19. [27]

    In: Pro- ceedings of the 2016 conference on empirical methods in natu ral language process- ing

    Taghipour, K., Ng, H.T.: A neural approach to automated e ssay scoring. In: Pro- ceedings of the 2016 conference on empirical methods in natu ral language process- ing. pp. 1882–1891 (2016)

  20. [28]

    In: Proceedings of the 16th Conference of t he European Chapter of the Association for Computational Linguistics: Main Vol ume

    de Vassimon Manela, D., Errington, D., Fisher, T., van Br eugel, B., Minervini, P.: Stereotype and skew: Quantifying gender bias in pre-tra ined and fine-tuned language models. In: Proceedings of the 16th Conference of t he European Chapter of the Association for Computational ...

  21. [29]

    arXiv preprint arXiv:2203.11171 (2022)

    Wang, X., Wei, J., Schuurmans, D., Le, Q., Chi, E., Narang , S., Chowdhery, A., Zhou, D.: Self-consistency improves chain of thought reaso ning in language models. arXiv preprint arXiv:2203.11171 (2022)

  22. [30]

    Educational measurement: issues and prac tice 31(1), 2–13 (2012)

    Williamson, D.M., Xi, X., Breyer, F.J.: A framework for e valuation and use of au- tomated scoring. Educational measurement: issues and prac tice 31(1), 2–13 (2012)

  23. [31]

    In: Proceedings of the 15th International Learning Analytics and Knowledge Conference

    Xiao, C., Ma, W., Song, Q., Xu, S.X., Zhang, K., Wang, Y., F u, Q.: Human-ai collaborative essay scoring: A dual-process framework wit h llms. In: Proceedings of the 15th International Learning Analytics and Knowledge Conference. pp. 293– 305 (2025) Large Language Model Bias i...

  24. [32]

    Nature Human Be haviour 8(10), 1839– 1850 (2024)

    Yan, L., Greiff, S., Teuber, Z., Gašević, D.: Promises and challenges of generative artificial intelligence for human learning. Nature Human Be haviour 8(10), 1839– 1850 (2024)

  25. [33]

    In: Proceedings of the 18th Workshop o n Innovative Use of NLP for Building Educational Applications (BEA 2023)

    Yancey, K.P., Laflair, G., Verardi, A., Burstein, J.: Rat ing short l2 essays on the cefr scale with gpt-4. In: Proceedings of the 18th Workshop o n Innovative Use of NLP for Building Educational Applications (BEA 2023). pp. 5 76–584 (2023)

  26. [34]

    In: Proceedings of the AAA I Conference on Artificial Intelligence

    Yang, K., Raković, M., Li, Y., Guan, Q., Gašević, D., Chen , G.: Unveiling the tapestry of automated essay scoring: A comprehensive inves tigation of accuracy, fairness, and generalizability. In: Proceedings of the AAA I Conference on Artificial Intelligence. vol. 38, pp. 22466...

  27. [35]

    In: Findings of the Association for Computational Linguis- tics: EMNLP 2020

    Yang, R., Cao, J., Wen, Z., Wu, Y., He, X.: Enhancing autom ated essay scor- ing performance via fine-tuning pre-trained language model s with combination of regression and ranking. In: Findings of the Association for Computational Linguis- tics: EMNLP 2020. pp. 1560–1569 (2020)

  28. [36]

    Reading and Writing 34(3), 705–725 (2021)

    Yoon, H.J.: Interactions in efl argumentative writing: E ffects of topic, l1 back- ground, and l2 proficiency on interactional metadiscourse. Reading and Writing 34(3), 705–725 (2021)

  29. [37]

    In: International Conferen ce on Artificial Intelli- gence in Education

    Yoshida, L.: The impact of example selection in few-shot prompting on automated essay scoring using gpt models. In: International Conferen ce on Artificial Intelli- gence in Education. pp. 61–73. Springer (2024)

  30. [38]

    In: Proceedings of the tenth workshop o n innovative use of NLP for building educational applications

    Zesch, T., Wojatzki, M., Scholten-Akoun, D.: Task-inde pendent features for auto- mated essay grading. In: Proceedings of the tenth workshop o n innovative use of NLP for building educational applications. pp. 224–232 (20 15)

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.