EXPLORING HUMAN FACTORS IN LARGE LANGUAGE MODEL-ASSISTED SOFTWARE REQUIREMENTS SPECIFICATION

Date

2026-4-3

Author

Kocakaya, Başak Düşün

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

51
views

0
downloads

Cite This

Large Language Models (LLMs) are being integrated into multiple phases of the software development life cycle, from requirements specification to maintenance. LLMs' technical benefits on software development, such as accelerating code generation and improving defect detection have become major breakthroughs in traditional software engineering workflows. However, the impact of using these models professionally on human aspects has not been comprehensively acknowledged, specifically in requirements specification. Therefore, this thesis examined how LLM-assisted software requirements specification impacts four key human aspects of software practitioners: perceived task complexity, task motivation, sense of achievement, and creative self-efficacy. The impact was examined through a within-subject experiment conducted with 30 software practitioners, where each of them completed a software requirements specification task without and with LLM assistance. The impact was examined through a within-subject experiment conducted with 30 software practitioners, who were required to write software requirements for a scenario under two conditions: without and with LLM assistance. After the task execution sessions, participants filled out validated scales on the four human aspects and participated in semi-structured interviews to discuss both their experiences and their responses to the scales. This research provides empirical evidence on how LLM-assisted software requirements specification shapes human aspects and underscores the need to account for these influences when integrating LLM tools into requirement engineering workflows.

Subject Keywords

LLM, Software Engineering, Requirements Specification, Task Complexity, Motivation, Sense of Achievement, Creative Self-Efficacy

URI

https://hdl.handle.net/11511/119103

Collections

Graduate School of Informatics, Thesis