A Multi-Dialectal, Longitudinal Corpus of Human-AI Hybrid Written Production

Full Official Name: A Multi-Dialectal, Longitudinal Corpus of Human-AI Hybrid Written Production
Submission date: Oct. 16, 2025, 5:52 a.m.

A multi-dialectal, longitudinal corpus of human-AI hybrid written production, including human-only, AI-only, and hybrid texts produced under various conditions. The corpus contains contributions from 693 speakers across five English national dialects, American English, Canadian English, British English, Australian English, and New Zealand English, with natural and hybrid samples paired for the same individuals over a four-week period. This design allows investigation of the short- and longer-term effects of LLM assistance on language use across geographic groups. The resource supports research on human-AI interaction, dialectal variation, and longitudinal changes in written production.

Creator(s)
University of Canterbury - Qiao Gan
University of Illinois Urbana-Champaign - Jonathan Dunn
University of Canterbury - Benjamin Adams
University of Manchester - Andrea Nini
Distributor(s)
University of Canterbury - Qiao Gan
University of Canterbury - Benjamin Adams
University of Illinois Urbana-Champaign - Jonathan Dunn
University of Manchester - Andrea Nini
Right Holder(s)
University of Canterbury - Qiao Gan
University of Canterbury - Benjamin Adams
University of Illinois Urbana-Champaign - Jonathan Dunn
University of Manchester - Andrea Nini
Status : Accepted
ISLRN :
119-185-989-055-2
Version
1.0
Source
https://github.com/GanQiaoUC/LREC_MulDiLoC-HAP
Resource Type
Primary Text
Media Type
Text
Language(s)
English
Access Medium
-