GenAI Embedded in Activities of Academic Research: Experiences and Lessons from HCI Studies

Abstract

Introduction: Generative Artificial Intelligence (GenAI) models, particularly ChatGPT, are increasingly being used to support various stages of scientific research. Objective: This paper aims to report on nine specific instances of the author team’s experience using ChatGPT to support their research activities within the field of Human-Computer Interaction (HCI). The goal is to identify and present lessons learned from these applications. Methodology: The methodology involved documenting and analyzing nine distinct cases where the author team utilized ChatGPT to assist with different research tasks. The analysis focused on identifying patterns of effective use and drawing conclusions regarding the tool’s strengths and limitations. Results: The analysis of these nine cases yielded six key lessons, demonstrating that ChatGPT provides substantial support across several research tasks. These tasks include content refinement, text reduction, insight generation, screening, summarization, and data visualization. The primary benefits observed were accelerated workfows, improved clarity, and enhanced systematic organization. However, the study also revealed that the effectiveness of ChatGPT is highly dependent on the use of precise, well-structured prompts and the segmentation of complex tasks into smaller, manageable units.

Ähnliche Arbeiten