AI partner versus human partner: comparing AI-based peer assessment with human-generated peer assessment in examining writing skills

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F62690094%3A18450%2F25%3A50022495" target="_blank" >RIV/62690094:18450/25:50022495 - isvavai.cz</a>

Result on the web

<a href="https://languagetestingasia.springeropen.com/articles/10.1186/s40468-025-00383-8" target="_blank" >https://languagetestingasia.springeropen.com/articles/10.1186/s40468-025-00383-8</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1186/s40468-025-00375-8" target="_blank" >10.1186/s40468-025-00375-8</a>

Result language

angličtina

Original language name

AI partner versus human partner: comparing AI-based peer assessment with human-generated peer assessment in examining writing skills

Original language description

This paper delves into the critical role of feedback in students’ peer assessment, highlighting its variability across different educational settings and its profound potential to enhance learning outcomes, particularly within traditional and AI learning environments. It explores the evolution of feedback influenced by technological advancements such as AI, focusing on their application to improve EFL college students’ writing skills through peer assessment. Using a rigorous a qualitative-dominant mixed method approach by integrating both quantitative and qualitative data, the study contrasts the effectiveness of traditional peer assessment (group A) with AI-based assessment using ChatGPT (group B) among fourth-year students from two different contexts, namely the University of Diyala, Iraq, and the University of Hradec Kralove, Czech Republic. While group A shows consistent improvement in writing skills, group B demonstrates slightly lower scores but offers quicker, accurate, and more precise feedback. The results of the study reveal significant differences between group A, utilizing traditional peer assessment, and group B, employing ChatGPT for AI-based assessment. In both contexts, group A’s students demonstrated consistent improvement in writing skills, with final scores ranging from 7 to 14 out of 15. Group B also showed improvement, albeit slightly less pronounced, with scores ranging from 6 to 12 out of 15. Teachers’ evaluations indicated that while group A benefited from reciprocal learning processes (writing and assessment), and greater social and cognitive engagement, group B experienced more accurate, quicker, and comprehensive feedback during peer assessment process from AI, albeit with less emotional and cognitive engagement. Ultimately, both methods contributed positively to students’ writing skills, highlighting the strengths and trade-offs between human and AI feedback mechanisms.Correction (publisher&apos;s corrections - author names and affiliations: https://www.webofscience.com/wos/woscc/full-record/WOS:001528974100001 ; Accession NumberWOS:001528974100001

Czech name

—

Czech description

—

Type

J_imp - Article in a specialist periodical, which is included in the Web of Science database

CEP classification

—

OECD FORD branch

60203 - Linguistics

Project

—

Continuities

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Publication year

2025

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Name of the periodical

Language Testing in Asia

ISSN

2229-0443

e-ISSN

2229-0443

Volume of the periodical

15

Issue of the periodical within the volume

1

Country of publishing house

GB - UNITED KINGDOM

Number of pages

30

Pages from-to

"Article Number: 38"

UT code for WoS article

001528974100001

EID of the result in the Scopus database

2-s2.0-105010229254