Verification of the Utility and Validity of Counseling AI Using Large Language Models

Verification of the Utility and Validity of Counseling AI Using Large Language Models - Verification of the Utility and Validity of Counseling AI Using Large Language Models

Status

Recruiting

Phases

Unknown

Study type

Observational

Source

JPRN

Registry ID

JPRN-UMIN000055643

Enrollment

Registered

2024-09-27

Start date

2024-01-17

Completion date

Unknown

Last updated

2026-06-29

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

Healthy individuals

Interventions

None listed

Eligibility

Sex/Gender

All

Inclusion criteria

Inclusion criteria: Individuals aged 20 to 65 at the time of obtaining consent. Individuals capable of providing informed consent after understanding the details of this study. Individuals with access to an internet environment that allows online counseling (use of free WiFi is not permitted) and who own both a computer and a smartphone. Individuals engaged in psychological counseling; those who are certified public psychologists or licensed clinical psychologists are preferred. University students or graduate students undergoing training in counseling programs are also eligible. Individuals not undergoing treatment for any mental disorder at the time of participation in the study.

Exclusion criteria

Exclusion criteria: Individuals with intellectual disabilities, dementia, or any language or thought disorders that may hinder communication during consent acquisition or counseling. Individuals whose native language is not Japanese. Any other individuals deemed ineligible for participation by the researchers. Individuals currently receiving counseling or psychiatric treatment at the time of study participation.

Design outcomes

Primary

Measure	Time frame
Effectiveness of Utterances Generated by Counseling AI: Assessing whether the AI is effective in functioning as a counselor. Validity of Utterances Generated by Counseling AI: Determining if the AI can generate expected responses and maintain coherent conversations. Comparison Between AI-Generated and Human-Generated Utterances: Evaluating whether the AI generates responses similar to those of highly-rated human counselors.	—

Measure

Time frame

Effectiveness of Utterances Generated by Counseling AI: Assessing whether the AI is effective in functioning as a counselor. Validity of Utterances Generated by Counseling AI: Determining if the AI can generate expected responses and maintain coherent conversations. Comparison Between AI-Generated and Human-Generated Utterances: Evaluating whether the AI generates responses similar to those of highly-rated human counselors.

—

Secondary

Measure	Time frame
Relationship with User Characteristics and Preferences: Examining whether user characteristics and preferences (such as gender, age, and counseling needs) are related to the evaluation of AI-generated utterances.	—

Countries

Japan

Contacts

Public ContactZui Narita

National Center of Neurology and Psychiatry National Institute of Mental Health

zuinarita@ncnp.go.jp042-341-2711

Outcome results