paperbot · PL 论文追踪

全部论文 待读 收藏

P2 Trust Dynamics in AI-Assisted Development: Definitions, Factors, and Implications

AI 辅助开发中的信任动态:定义、影响因素与启示
ICSE 2025 IEEE/ACM 47th International Conference on Software Engineering (ICSE)2025匹配度 2/10 · 该论文属于开发者与 AI 代码助手协作的人因/HCI 实证研究,关注信任定义与交互指南,不涉及静态分析、形式化验证、浮点程序分析或贝叶斯程序分析等核心方向,与用户研究画像交集很小。引用 5 inbox
Sadra Sabouri, Philipp Eibl, Xinyi Zhou, Morteza Ziyadi, Nenad Medvidovic, Lars Lindemann, Souti Chattopadhyay
crossref: 2026-09-12

AI 速览

一句话:通过混合方法研究开发者如何定义、评估并动态调整对 AI 代码建议的信任,提炼并验证四条交互指南。

该研究关注开发者如何判断是否该信任 AI 代码生成建议,因为接受“好”代码、拒绝“坏”代码对开发者-AI 协作编程至关重要。作者采用混合方法:先对 29 名开发者进行探索性问卷调查,再对 10 名开发者进行观察研究,分析开发者(1)如何定义代码建议的可信度,(2)如何评估可信度,(3)信任在使用 AI 助手过程中如何演变。结果发现,可理解性与感知正确性是最常用的评估因素;但开发者对信任的定义与评估之间存在缺口,说明实时评估可信代码缺乏支持。研究还发现开发者经常改变信任决策,仅保留 52% 的原始建议。基于此,作者提炼出四条增强开发者-AI 交互的指南,并通过 7 名领域专家和 8 名问卷成员的调查进行验证,同时讨论了指南的应用方式与工具支持。

问题
开发者需要知道何时该信任 AI 代码建议,但开发者如何定义、评估可信度以及信任如何随时间演变尚不清晰。
方法
采用混合方法研究,包括对 29 名开发者的深度探索性问卷和对 10 名开发者的观察研究,并基于发现提炼指南后用专家与开发者调查进行验证。
评估
通过问卷与观察数据归纳可信度评估因素和信任变化模式,再以 7 名领域专家和 8 名问卷成员的调查验证所提出的四条交互指南。
原文摘要(Abstract)

Software developers increasingly rely on AI code generation utilities. To ensure that “good” code is accepted into the code base and “bad” code is rejected, developers must know when to trust an AI suggestion. Understanding how developers build this intuition is crucial to enhancing developer-AI collaborative programming. In this paper, we seek to understand how developers (1) define and (2) evaluate the trustworthiness of a code suggestion and (3) how trust evolves when using AI code assistants. To answer these questions, we conducted a mixed method study consisting of an in-depth exploratory survey with (n=29) developers followed by an observation study (n=10). We found that comprehensibility and perceived correctness were the most frequently used factors to evaluate code suggestion trustworthiness. However, the gap in developers' definition and evaluation of trust points to a lack of support for evaluating trustworthy code in real-time. We also found that developers often alter their trust decisions, keeping only 52% of original suggestions. Based on these findings, we extracted four guidelines to enhance developer-AI interactions. We validated the guidelines through a survey with (n=7) domain experts and survey members (n=8). We discuss the validated guidelines, how to apply them, and tools to help adopt them.

链接与引用

DOI 原文 ·

BibTeX
@inproceedings{paperbot8855,
  title = {Trust Dynamics in AI-Assisted Development: Definitions, Factors, and Implications},
  author = {Sadra Sabouri and Philipp Eibl and Xinyi Zhou and Morteza Ziyadi and Nenad Medvidovic and Lars Lindemann and Souti Chattopadhyay},
  booktitle = {ICSE},
  year = {2025},
  doi = {10.1109/icse55347.2025.00199}
}