Saved in:
Bibliographic Details
Main Authors: Hao, Jiangang, Cui, Wenju, Kyllonen, Patrick, Kerzabi, Emily, Liu, Lei, Flor, Michael
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2411.10246
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918366873124864
author Hao, Jiangang
Cui, Wenju
Kyllonen, Patrick
Kerzabi, Emily
Liu, Lei
Flor, Michael
author_facet Hao, Jiangang
Cui, Wenju
Kyllonen, Patrick
Kerzabi, Emily
Liu, Lei
Flor, Michael
contents Collaborative problem solving (CPS) is widely recognized as a critical 21st-century skill. Assessing CPS depends heavily on coding the communication data using a construct-relevant framework, and this process has long been a major bottleneck to scaling up such assessments. Based on five datasets and two coding frameworks, we demonstrate that ChatGPT can code communication data to a satisfactory level, though performance varies across ChatGPT models, and depends on the coding framework and task characteristics. Interestingly, newer reasoning-focused models such as GPT-o1-mini and GPT-o3-mini do not necessarily yield better coding results. Additionally, we show that refining prompts based on feedback from miscoded cases can improve coding accuracy in some instances, though the effectiveness of this approach is not consistent across all tasks. These findings offer practical guidance for researchers and practitioners in developing scalable, efficient methods to analyze communication data in support of 21st-century skill assessment.
format Preprint
id arxiv_https___arxiv_org_abs_2411_10246
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Automated Coding of Communications in Collaborative Problem-solving Tasks Using ChatGPT
Hao, Jiangang
Cui, Wenju
Kyllonen, Patrick
Kerzabi, Emily
Liu, Lei
Flor, Michael
Human-Computer Interaction
Computation and Language
Collaborative problem solving (CPS) is widely recognized as a critical 21st-century skill. Assessing CPS depends heavily on coding the communication data using a construct-relevant framework, and this process has long been a major bottleneck to scaling up such assessments. Based on five datasets and two coding frameworks, we demonstrate that ChatGPT can code communication data to a satisfactory level, though performance varies across ChatGPT models, and depends on the coding framework and task characteristics. Interestingly, newer reasoning-focused models such as GPT-o1-mini and GPT-o3-mini do not necessarily yield better coding results. Additionally, we show that refining prompts based on feedback from miscoded cases can improve coding accuracy in some instances, though the effectiveness of this approach is not consistent across all tasks. These findings offer practical guidance for researchers and practitioners in developing scalable, efficient methods to analyze communication data in support of 21st-century skill assessment.
title Automated Coding of Communications in Collaborative Problem-solving Tasks Using ChatGPT
topic Human-Computer Interaction
Computation and Language
url https://arxiv.org/abs/2411.10246