Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2508.00847 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913971239387136 |
|---|---|
| author | Sahab, Sofia Haqbeen, Jawad Sapkota, Diksha Ito, Takayuki |
| author_facet | Sahab, Sofia Haqbeen, Jawad Sapkota, Diksha Ito, Takayuki |
| contents | In this study, we investigated the effects of GPT-4, with and without specific conversational instructions, on the mental health of Afghan women. These women face multifaceted challenges, including Taliban-imposed restrictions, societal inequalities, and domestic violence, adversely affecting their well-being. We conducted a randomized controlled trial with 60 participants, dividing them into three groups: GPT-4, a supportive listener (GPT-4 with empathetic engagement instructions), and a waiting list. The Hospital Anxiety and Depression Scale (HADS) was used to measure anxiety and depression before and after the intervention. Linguistic analysis of chat data examined personal pronouns, tones, emotions, and Language Style Matching (LSM). The supportive listener group showed a significant reduction in HADS scores compared to the other groups. Linguistic analysis revealed a more positive tone and higher LSM in the supportive listener group, with a significant negative correlation between LSM and changes in HADS scores, indicating greater linguistic alignment was linked to reductions in anxiety and depression. Perceived empathy ratings were also significantly higher in the supportive listener group. These findings highlight the potential of AI-driven interventions, like GPT-4, in providing accessible mental health support. However, such interventions should complement traditional psychotherapy, ensuring a collaborative approach to optimize therapeutic outcomes. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_00847 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | GPT Chatbots for Alleviating Anxiety and Depression: A Pilot Randomized Controlled Trial with Afghan Women Sahab, Sofia Haqbeen, Jawad Sapkota, Diksha Ito, Takayuki Human-Computer Interaction Computers and Society In this study, we investigated the effects of GPT-4, with and without specific conversational instructions, on the mental health of Afghan women. These women face multifaceted challenges, including Taliban-imposed restrictions, societal inequalities, and domestic violence, adversely affecting their well-being. We conducted a randomized controlled trial with 60 participants, dividing them into three groups: GPT-4, a supportive listener (GPT-4 with empathetic engagement instructions), and a waiting list. The Hospital Anxiety and Depression Scale (HADS) was used to measure anxiety and depression before and after the intervention. Linguistic analysis of chat data examined personal pronouns, tones, emotions, and Language Style Matching (LSM). The supportive listener group showed a significant reduction in HADS scores compared to the other groups. Linguistic analysis revealed a more positive tone and higher LSM in the supportive listener group, with a significant negative correlation between LSM and changes in HADS scores, indicating greater linguistic alignment was linked to reductions in anxiety and depression. Perceived empathy ratings were also significantly higher in the supportive listener group. These findings highlight the potential of AI-driven interventions, like GPT-4, in providing accessible mental health support. However, such interventions should complement traditional psychotherapy, ensuring a collaborative approach to optimize therapeutic outcomes. |
| title | GPT Chatbots for Alleviating Anxiety and Depression: A Pilot Randomized Controlled Trial with Afghan Women |
| topic | Human-Computer Interaction Computers and Society |
| url | https://arxiv.org/abs/2508.00847 |