人工智能可解释性信息对专业心理求助的影响：基于认知与信息的双路径整合框架

doi:10.3724/SP.J.1042.2026.LS.00142

• •

人工智能可解释性信息对专业心理求助的影响：基于认知与信息的双路径整合框架

马诗浩, 叶建梅, 陈倩, 王伟军

遵义师范学院教师教育学院, 贵州 563006 中国
华中师范大学青少年网络心理与行为教育部重点实验室, 湖北 430079 中国
华中师范大学心理学院, 湖北 430079 中国
北京城市学院绿萝社会心理智能创新发展研究院, 北京 100094 中国

收稿日期:2026-07-02 修回日期:2026-09-17 接受日期:2026-10-10
基金资助:
人本人工智能驱动的信息服务体系重构与应用研究(22&ZD324); 可解释人工智能影响青少年专业心理求助行为的作用机制(遵师BS[2026]12号)

Understanding the Influence of AI Explainability Information on Professional Mental Health Help-Seeking: An Integrated Cognitive–Information Dual-Pathway Framework

Ma Shihao, Ye Jianmei, Chen Qian, Wang Weijun

School of Teacher Education, Zunyi Normal University 563006, China
Key Laboratory of Adolescent Cyberpsychology and Behavior(CCNU), Central China Normal University 430079, China
School of Psychology, Central China Normal University 430079, China
BeiCheng Lvluo Institute for Social Psychological Intelligence Innovation and Development, Beijing City University 100094, China

Received:2026-07-02 Revised:2026-09-17 Accepted:2026-10-10

摘要/Abstract

摘要： 随着生成式人工智能(Generative AI)在心理健康领域的广泛应用，人工智能(AI)可解释性逐渐成为影响个体健康决策的关键因素。然而，可解释性的现有研究多聚焦技术层面，较少从个体认知与信息加工视角探讨人工智能提供的解释性信息如何影响专业心理求助。本文以双加工理论为总体框架，结合健康信念模型(HBM)、扩展平行过程模型(EPPM)以及信息寻求整合模型(CMIS)，从认知加工和解释评估两个层面构建人工智能可解释性信息影响专业心理求助的双路径整合框架。该框架提出，解释性信息通过两条路径发挥作用：一是解释性信息内容(如健康状态信息、求助资源信息、注意偏好信息)经由心理问题认知的认知路径；二是解释性信息特征(如解释性方法、量化数据、信息来源)经由解释性评估的启发式路径。两条路径既独立发挥作用，又存在链式中介效应。此外，知识储备、心理问题类型与情境因素以及个体因素在以上路径中起调节作用。该框架为人工智能可解释性在心理健康服务领域的应用提供了理论指导，也为后续实证研究奠定了可检验的理论基础。

关键词: 人工智能可解释性, 专业心理求助, 双路径整合框架

Abstract: With the increasing application of generative artificial intelligence (AI) in the field of mental health, AI explainability has emerged as a critical factor shaping individuals’ health-related decision-making. However, existing research on explainability has predominantly focused on technological aspects, with limited attention to how AI-generated explanatory information influences professional psychological help-seeking behavior from the perspectives of individual cognition and information processing. Drawing upon dual-process theories as the overarching framework, this article integrates the Health Belief Model (HBM), the Extended Parallel Process Model (EPPM), and the Comprehensive Model of Information Seeking (CMIS) to propose a dual-pathway integrative framework explaining how AI explainability information may influence professional psychological help-seeking behavior through cognitive processing and explanatory evaluation mechanisms.Specifically, the proposed framework suggests that explainability information may operate through two complementary pathways. The first is a cognitive pathway, in which the content of explainability information (e.g., health status information, help-seeking resource information, and attentional preference information) facilitates psychological problem recognition. The second is a heuristic pathway, in which the features of explainability information (e.g., explanation methods, quantitative data presentation, and information sources) shape individuals’ explanatory evaluations. These two pathways may function independently while also exhibiting potential sequential mediation effects. Furthermore, individual knowledge resources, types of psychological problems, contextual characteristics, and personal factors may moderate these pathways.This framework provides theoretical insights into the application of AI explainability in mental health services and establishes a theoretically testable foundation for future empirical investigations.

Key words: AI explainabilit, professional mental health help-seeking, dual-pathway integrative framework

马诗浩, 叶建梅, 陈倩, 王伟军. 人工智能可解释性信息对专业心理求助的影响：基于认知与信息的双路径整合框架. , doi: 10.3724/SP.J.1042.2026.LS.00142.

Ma Shihao, Ye Jianmei, Chen Qian, Wang Weijun. Understanding the Influence of AI Explainability Information on Professional Mental Health Help-Seeking: An Integrated Cognitive–Information Dual-Pathway Framework. , doi: 10.3724/SP.J.1042.2026.LS.00142.