Validity Assessment and Ethical Challenges of Using Large Language Models (LLMs) to Identify Schema-Related Patterns in Clinical Psychotherapy: A Structured Literature Review
Validity Assessment and Ethical Challenges of Using Large Language Models (LLMs) to Identify Schema-Related Patterns in Clinical Psychotherapy: A Structured Literature Review
jafar askarishahid1
1) M.A. Student in General Psychology, Islamic Azad University, Isfahan (Khorasgan) Branch Email:
IOI Code
:
ACZF-ACAGDC
Publication :
6th.International Congress on Management, Economy, Humanities, Education and Business Development(6icmba.ir)
Abstract :
Large language models (LLMs), through their ability to process and analyze textual data, have created new opportunities for psychological assessment and the identification of cognitive and emotional patterns. One potential application of these technologies is the identification of schema-related patterns in clinical psychotherapy. However, the validity of inferences generated by these models and the ethical implications of their use remain challenging issues. The present study aimed to examine the potential capabilities, validity limitations, and ethical challenges of using LLMs to identify schema-related patterns. A structured literature review with a narrative synthesis approach was conducted using sources retrieved from Google Scholar, PubMed, and APA PsycINFO. Relevant studies addressing the applications of artificial intelligence and language models in clinical psychology, language analysis, psychological assessment, and ethical considerations were reviewed and thematically categorized. The findings suggest that LLMs have the potential to identify linguistic, cognitive, and emotional patterns associated with schemas. Nevertheless, extending these capabilities to the valid identification of early maladaptive schemas requires independent psychometric and clinical validation. Furthermore, limited interpretability, algorithmic bias, cultural and linguistic differences, and risks to the confidentiality of clients data represent major challenges to the application of these technologies. Accordingly, LLMs should be used as supportive tools for psychological assessment and clinical decision-making, under professional supervision and subject to independent validation, rather than as substitutes for clinicians professional judgment.
Keywords : large language models (llms) cognitive schemas clinical validity digital psychotherapy ai ethics