Paper Number
ECIS2026-1777
Paper Type
CRP
Abstract
This study examines how personality congruence between humans and large language models (LLMs) influences creative problem-solving. Building on the Human–AI Collaboration framework and Social Presence Theory, the research investigates how alignment between user personality (introversion or extraversion) and LLM interaction style affects creativity, engagement, and task efficiency. Using the HEXACO-PI-R model to classify participants, matched (introvert–introvert, extrovert–extrovert) and mismatched (introvert–extrovert, extrovert–introvert) human–LLM pairs were tested in controlled experiments with GPT-based agents trained for specific personality expressions. Findings are expected to reveal how personality fit enhances perceived social presence, reduces cognitive effort, and improves creative outcomes measured by innovativeness, usefulness, and impact. The study advances theory by integrating personality psychology with AI interaction design, offering empirical evidence on the behavioural mechanisms underlying human–LLM collaboration. Practically, it provides insights for designing adaptive, personality-aware LLM systems that foster effective and meaningful creative partnerships.
Recommended Citation
Jha, Ashish Kumar and Jasim, K. Mohamed, "Enhancing Human-Ai Collaboration: An Experimental Study On Ai-Human Collaborative Productivity" (2026). ECIS 2026 Proceedings. 12.
https://aisel.aisnet.org/ecis2026/gen_track/gen_track/12
Enhancing Human-Ai Collaboration: An Experimental Study On Ai-Human Collaborative Productivity
This study examines how personality congruence between humans and large language models (LLMs) influences creative problem-solving. Building on the Human–AI Collaboration framework and Social Presence Theory, the research investigates how alignment between user personality (introversion or extraversion) and LLM interaction style affects creativity, engagement, and task efficiency. Using the HEXACO-PI-R model to classify participants, matched (introvert–introvert, extrovert–extrovert) and mismatched (introvert–extrovert, extrovert–introvert) human–LLM pairs were tested in controlled experiments with GPT-based agents trained for specific personality expressions. Findings are expected to reveal how personality fit enhances perceived social presence, reduces cognitive effort, and improves creative outcomes measured by innovativeness, usefulness, and impact. The study advances theory by integrating personality psychology with AI interaction design, offering empirical evidence on the behavioural mechanisms underlying human–LLM collaboration. Practically, it provides insights for designing adaptive, personality-aware LLM systems that foster effective and meaningful creative partnerships.