Rational Emotive Behavior Therapy
Cognitive Therapy
Stress Prevention and Stress Management Techniques VI
Modeling in Therapy
Interpersonal Psychotherapy
Introspection
You might also read
Articles linked to this work by shared authors, journal, and citation graph.
Updated: Sep 22, 2025

Virtual Prism Adaptation Therapy: Protocol for Validation in Healthy Adults
Published on: February 12, 2020
Navot Naor1, Alex Frenkel1, Mirène Winsberg2
1Kai.ai, Tel Aviv, Israel.
This study examined whether a mobile artificial intelligence tool, Kai.ai, could improve user well-being when delivered through common messaging platforms. By analyzing data from over 2,900 users, researchers found that participants reported higher well-being scores after using the service. The results suggest that AI-powered psychological support delivered via text messaging is a promising approach for enhancing mental health.
Area of Science:
Background:
No prior work had fully resolved the efficacy of artificial intelligence-driven psychological support delivered through standard messaging platforms. Smartphone applications have recently seen a massive increase in popularity for delivering behavioral health coaching. Acceptance Commitment Therapy protocols are established as effective treatments for anxiety and depression symptoms when provided via mobile software. That uncertainty drove the need to investigate whether similar benefits persist when using automated conversational agents. Prior research has shown that digital interventions offer scalable solutions for mental health support. However, the specific impact of AI-powered tools on user well-being remains an area requiring rigorous evaluation. This gap motivated the current retrospective analysis of a widely accessible digital intervention. Researchers aimed to determine if these automated systems could reliably improve psychological outcomes for a broad user base.
Purpose Of The Study:
The aim of this study is to expand on existing research and test the suitability of artificial intelligence-driven interventions delivered through popular texting applications. Researchers sought to determine if these automated systems could effectively improve user well-being. This investigation specifically evaluated the hypothesis that interacting with the Kai.ai platform results in measurable psychological benefits. The study addresses the need for scalable mental health support in an increasingly digital world. By leveraging common messaging platforms, the intervention aims to reach a wide audience with minimal barriers to access. The authors intended to quantify the impact of daily engagement on long-term well-being outcomes. This work builds upon previous evidence regarding the effectiveness of Acceptance Commitment Therapy protocols in mobile formats. The project provides a necessary assessment of how AI-powered tools function in real-world, non-clinical settings.
Main Methods:
The review approach involved a pragmatic retrospective analysis of 2,909 individuals who utilized the specified digital service. Investigators tracked participant progress through popular messaging applications including Discord and Telegram. Researchers employed the World Health Organization-Five Well-Being Index to monitor psychological states at multiple time points. A 1-tailed paired samples t test assessed differences between baseline and final scores. Hierarchical linear modeling examined how symptom trajectories shifted during the engagement period. This statistical framework allowed for the evaluation of longitudinal data patterns. The team analyzed the relationship between daily message frequency and reported well-being outcomes. This methodology ensured a comprehensive assessment of the intervention's impact across a diverse user group.
Main Results:
Key findings from the literature indicate that the median well-being score increased from 40 at baseline to 52 at the final measurement. This improvement reached statistical significance with a reported W-value of 2,682,927 and a P-value below 0.001. Hierarchical linear modeling revealed that well-being gains were linearly associated with the volume of daily messages sent by users. The beta coefficient for daily message frequency was 0.029, with a t-statistic of 4 and a P-value under 0.001. The interaction between the number of messages and the unique number of days also showed a significant effect. This interaction term yielded a beta coefficient of -0.0003, a t-statistic of -2.2, and a P-value of 0.03. These results suggest a consistent positive trend in user well-being throughout the intervention period. The data support the efficacy of the mobile-based approach in fostering psychological improvements.
Conclusions:
The authors propose that mobile-based Acceptance Commitment Therapy interventions serve as effective tools for enhancing individual well-being. Their analysis suggests that automated conversational agents can successfully support mental health through popular messaging platforms. Findings indicate that greater engagement with the service correlates with more substantial improvements in reported well-being scores. The researchers suggest that this specific AI-driven approach holds significant promise for maintaining high levels of psychological health. These results imply that text-based digital interventions provide a viable pathway for scalable mental health support. The study demonstrates that users experience measurable benefits when interacting with the platform over time. This synthesis highlights the potential for integrating automated psychological coaching into daily digital communication habits. Future implementation of such technologies may offer accessible assistance for individuals seeking to improve their daily lives.
The researchers propose that engagement with the AI tool, measured by daily message volume, correlates with improved well-being. Participants showed a median score increase from 40 to 52 on the World Health Organization-Five Well-Being Index, indicating a significant positive shift in their mental state.
The study utilized Kai.ai, an artificial intelligence-driven intervention designed to deliver psychological support. This tool operates directly within popular messaging platforms like WhatsApp, iMessage, and Telegram, allowing for seamless integration into the user's existing digital communication environment.
The authors state that the use of a 1-tailed paired samples t test was necessary to assess the directional improvement in well-being levels. This statistical approach allowed for a direct comparison between the initial baseline scores and the final measurements taken after service engagement.
Hierarchical linear modeling served as the primary data type for examining symptom changes over time. This method allowed the researchers to account for the nested structure of the data, specifically linking the frequency of daily interactions to the observed longitudinal improvements in well-being.
The researchers measured well-being using the World Health Organization-Five Well-Being Index. This standardized tool tracked psychological states throughout the duration of the service engagement, providing a consistent metric to evaluate the efficacy of the intervention across the entire user population.
The authors propose that their findings demonstrate the great promise of this AI-driven approach in helping individuals maintain high levels of well-being. They suggest that such mobile-based interventions are effective means to improve the daily lives of users who engage with the service.