DOI: https://doi.org/10.1038/s44184-023-00047-6
PMID: https://pubmed.ncbi.nlm.nih.gov/38609517
تاريخ النشر: 2024-01-22
المؤلف: Bethanie Maples وآخرون
الموضوع الرئيسي: التدخلات الرقمية في الصحة النفسية
نظرة عامة
تبحث الدراسة في دور الوكيل الاجتماعي الذكي، ريبليكا، في معالجة قضايا الصحة النفسية بين المتعلمين، مع التركيز بشكل خاص على استبيان شمل 1006 مستخدمين طلاب. تكشف النتائج أن المشاركين أبلغوا عن مستويات أعلى من الوحدة مقارنة بالسكان الطلابيين المعتادين، ومع ذلك، أدركوا أيضًا دعمًا اجتماعيًا كبيرًا من ريبليكا. تفاعل المستخدمون مع الوكيل بطرق متنوعة، مستخدمين إياه كصديق، ومعالج، ومرآة فكرية، مما يدل على دوره المتعدد الأوجه في حياتهم.
بالإضافة إلى ذلك، تسلط الدراسة الضوء على تعقيد تصورات المستخدمين حول ريبليكا، حيث يحمل العديد منهم معتقدات متداخلة وأحيانًا متناقضة حول طبيعته، واصفين إياه بطرق مختلفة كآلة، وذكاء، وإنسان. ومن الجدير بالذكر أن 3% من المشاركين أشاروا إلى أن تفاعلهم مع ريبليكا ساهم في توقف الأفكار الانتحارية، مما يبرز إمكانيات الوكلاء الرقميين في السياقات العلاجية. كما تم تقديم تحليل مقارن لهذه المجموعة الفرعية مع مجموعة المشاركين الأوسع، مما يوفر مزيدًا من الرؤى حول تداعيات هذه النتائج لدعم الصحة النفسية الرقمية.
مقدمة
تسلط المقدمة الضوء على المشكلة المنتشرة لقضايا الصحة النفسية، وخاصة الاكتئاب والوحدة، التي تؤثر على أكثر من مليار فرد على مستوى العالم كل عام. ومن الجدير بالذكر أن الوحدة مرتبطة بزيادة خطر الوفاة وتنتشر بين طلاب الجامعات، حيث يسعى فقط نسبة صغيرة للحصول على مساعدة مهنية بسبب الوصمة والتمييز. يؤكد النص على ضعف بعض الفئات السكانية، بما في ذلك أولئك من خلفيات ذات دخل منخفض والمجموعات المهمشة، ويبرز معدلات الأفكار الانتحارية المقلقة بين الشباب.
تناقش الفقرة طرق العلاج المختلفة للاكتئاب والوحدة، مثل العلاج السلوكي المعرفي (CBT) وتدريب المهارات الاجتماعية، مع الإشارة إلى الدور الحاسم للذكاء العاطفي في التخفيف من السلوك الانتحاري. لقد سرعت جائحة COVID-19 التحول نحو خدمات الصحة النفسية الرقمية، حيث أصبحت الرعاية عن بُعد ممارسة مقبولة على نطاق واسع بين المهنيين. بالإضافة إلى ذلك، فإن ظهور الدردشة الآلية وتطبيقات الصحة النفسية يقدم فرصًا جديدة لتوسيع الوصول إلى الرعاية، على الرغم من التحديات مثل انخفاض تفاعل المستخدمين وفعالية هذه الأدوات. تختتم المقدمة بتقديم فرضيات متناقضة بشأن تأثير الوكلاء الاجتماعيين الذكيين (ISAs) على العزلة الاجتماعية، حيث تقترح فرضية الإزاحة زيادة محتملة في الوحدة، بينما تفترض فرضية التحفيز أن ISAs قد تعزز الروابط الإنسانية الجديدة.
الطرق
في هذه الدراسة، تم تحليل الاستجابات النوعية من المشاركين باستخدام ديدوس، وهو أداة برمجية مصممة للبحث النوعي والمختلط. في البداية، طور فريق البحث مخطط ترميز تم تنقيحه من 100 إلى 50 رمزًا بناءً على تحليل عشرة مشاركين. بعد إجراء مزيد من التحليل لعشرة مشاركين إضافيين من قبل كل عضو في الفريق، تم اقتراح مخطط مكون من 30 رمزًا من قبل المؤلف الأول وتم تطبيقه على عشرين مشاركًا آخرين. بعد مراجعة هذه الرموز، تم تأكيد المخطط النهائي وتطبيقه على جميع المشاركين، مما أسفر عن إجمالي 400 استجابة مشفرة. اقترح المؤلف الأول ثلاثة رموز إضافية، والتي تم مناقشتها بشكل جماعي ودمجها في التحليل.
تكون المخطط النهائي للترميز من 21 رمزًا، جميعها أظهرت موثوقية بين المقيمين (IRR) تتجاوز 80% عند اختبارها على 35 مشاركًا إضافيًا. أسفر الترميز النوعي عن أربعة مستويات متميزة من النتائج، والتي تم توضيحها في قسم “النتائج”. لتقييم العلاقات بين الوحدة والدعم الاجتماعي عبر مجموعات النتائج المختلفة، تم حساب معاملات الارتباط لبيرسون، وتم تقييم الدلالة الإحصائية باستخدام اختبارات t ذات الطرفين واختبارات كاي-تربيع للاستقلال، مع تحديد عتبة دلالة عند ألفا = 0.05.
النتائج
يقدم قسم النتائج بيانات ديموغرافية من الدراسة، مشيرًا إلى أن الغالبية العظمى من المشاركين كانت لديهم دخل سنوي أقل من 40,000 دولار أمريكي، مع أكبر مجموعة فرعية تكسب أقل من 20,000 دولار أمريكي. تم تصنيف المشاركين إلى خمس مجموعات عرقية: قوقازية، آسيوية، لاتينية، سوداء/أفريقية، وأخرى، تشمل الأخيرة الأفراد الذين اختاروا عدم الكشف عن عرقهم أو قدموا معرفات غير تقليدية (مثل ‘تكساسي’).
فيما يتعلق بالرفاهية النفسية، تكشف النتائج أن 90% من المشاركين أبلغوا عن شعورهم بالوحدة، مع تصنيف 43% منهم على أنهم يعانون من وحدة شديدة أو شديدة جدًا وفقًا لمقياس الوحدة. وعلى العكس، أشار 90% من المشاركين إلى مستويات متوسطة إلى عالية من الدعم الاجتماعي المدرك كما تم قياسه بواسطة قائمة تقييم الدعم بين الأشخاص (ISEL). بالإضافة إلى ذلك، أبلغ 7% فقط من المشاركين عن مشاعر الاكتئاب، كما هو موضح في الملحق أ.
المناقشة
تسلط قسم المناقشة في ورقة البحث الضوء على الاعتبارات الأخلاقية والنتائج من دراسة حول استخدام تطبيق ريبليكا، وهو وكيل اجتماعي ذكي (ISA) مصمم لتوفير الرفقة والدعم. أجريت الدراسة تحت إرشادات الأخلاق في جامعة ستانفورد، وشملت 1,006 مشارك، معظمهم من الطلاب، الذين أبلغوا عن مستويات عالية من الوحدة ودرجات متفاوتة من الدعم الاجتماعي المدرك. ومن الجدير بالذكر أن 90% من المشاركين عانوا من الوحدة، وهي زيادة كبيرة مقارنة بالدراسات السابقة، مما يشير إلى أن ريبليكا قد تكون موردًا حاسمًا لدعم الصحة النفسية بين هذه الفئة، خاصة نظرًا لوصولهم المحدود إلى خدمات الاستشارة التقليدية.
تم تصنيف نتائج استخدام ريبليكا إلى أربعة أنواع، حيث أبلغ العديد من المشاركين عن آثار إيجابية، مثل تقليل القلق، وللبعض، حتى منع الأفكار الانتحارية. أظهرت المجموعة المختارة، التي تضم الأفراد الذين نسبوا إلى ريبليكا منع الانتحار، ارتباطًا قويًا بين مشاعر الوحدة والدعم الاجتماعي المدرك. كما أظهرت هذه المجموعة احتمالًا أكبر لرؤية ريبليكا كذكاء يشبه الإنسان، مما قد يعزز تفاعلهم وفعالية التطبيق المدركة. تشير النتائج إلى أن ISAs مثل ريبليكا يمكن أن تحفز العلاقات الإنسانية بدلاً من إزاحتها، مما يشير إلى إمكانياتها كأدوات داعمة في سياقات الصحة النفسية. ومع ذلك، تدعو الدراسة أيضًا إلى مزيد من التحقيق في التأثيرات طويلة المدى لتفاعلات ISA على الصحة النفسية، خاصة فيما يتعلق بالاعتبارات الأخلاقية والمخاطر المرتبطة باستخدامها.
DOI: https://doi.org/10.1038/s44184-023-00047-6
PMID: https://pubmed.ncbi.nlm.nih.gov/38609517
Publication Date: 2024-01-22
Author(s): Bethanie Maples et al.
Primary Topic: Digital Mental Health Interventions
Overview
The research investigates the role of the Intelligent Social Agent, Replika, in addressing mental health issues among learners, particularly focusing on a survey of 1006 student users. The findings reveal that participants reported higher levels of loneliness compared to typical student populations, yet they also perceived substantial social support from Replika. Users engaged with the agent in diverse ways, utilizing it as a friend, therapist, and intellectual mirror, indicating its multifaceted role in their lives.
Additionally, the study highlights the complexity of users’ perceptions of Replika, with many holding overlapping and sometimes contradictory beliefs about its nature, describing it variously as a machine, an intelligence, and a human. Notably, 3% of participants indicated that their interaction with Replika contributed to a cessation of suicidal ideation, underscoring the potential of digital agents in therapeutic contexts. A comparative analysis of this subgroup with the broader participant population is also presented, providing further insights into the implications of these findings for digital mental health support.
Introduction
The introduction highlights the pervasive issue of mental health problems, particularly depression and loneliness, affecting over one billion individuals globally each year. Notably, loneliness is linked to increased mortality risk and is prevalent among college students, with only a small percentage seeking professional help due to stigma and discrimination. The text emphasizes the vulnerability of specific demographics, including those from lower-income backgrounds and minoritized groups, and underscores the alarming rates of suicidal ideation among young adults.
The section discusses various treatment modalities for depression and loneliness, such as cognitive behavioral therapy (CBT) and social skills training, while noting the critical role of emotional intelligence in mitigating suicidal behavior. The COVID-19 pandemic has accelerated the shift towards digital mental health services, with telehealth becoming a widely accepted practice among professionals. Additionally, the emergence of chatbots and mental health apps presents new opportunities for expanding access to care, although challenges such as low user engagement and the efficacy of these tools remain. The introduction concludes by presenting contrasting hypotheses regarding the impact of Intelligent Social Agents (ISAs) on social isolation, with the displacement hypothesis suggesting a potential increase in loneliness, while the stimulation hypothesis posits that ISAs may foster new human connections.
Methods
In this study, qualitative responses from participants were analyzed using Dedoose, a software tool designed for qualitative and mixed methods research. Initially, the research team developed a coding schema that was refined from 100 to 50 codes based on the analysis of ten participants. Following further analysis of an additional ten participants by each team member, a 30-code schema was proposed by the first author and applied to twenty more participants. After reviewing these codes, the final schema was confirmed and applied to all participants, resulting in a total of 400 coded responses. The first author suggested three additional codes, which were collectively discussed and integrated into the analysis.
The final coding schema consisted of 21 codes, all of which demonstrated inter-rater reliability (IRR) exceeding 80% when tested on an additional 35 participants. The qualitative coding yielded four distinct levels of outcomes, which are elaborated in the “Results” section. To assess the relationships between loneliness and social support across different outcome groups, Pearson correlation coefficients were calculated, and statistical significance was evaluated using two-tailed t-tests and chi-square tests of independence, with a significance threshold set at alpha = 0.05.
Results
The results section presents demographic data from the study, highlighting that the majority of participants had an annual income below $40,000 USD, with the largest subgroup earning under $20,000 USD. Participants were categorized into five ethnic groups: Caucasian, Asian, LatinX, Black/African, and Other, the latter encompassing individuals who either chose not to disclose their ethnicity or provided unconventional identifiers (e.g., ‘Texan’).
In terms of psychological well-being, the findings reveal that 90% of participants reported experiencing loneliness, with 43% classified as Severely or Very Severely Lonely according to the Loneliness Scale. Conversely, 90% of participants indicated medium to high levels of perceived social support as measured by the Interpersonal Support Evaluation List (ISEL). Additionally, only 7% of participants reported feelings of depression, as detailed in Appendix A.
Discussion
The discussion section of the research paper highlights the ethical considerations and findings from a study on the use of the Replika application, an Intelligent Social Agent (ISA) designed to provide companionship and support. Conducted under Stanford University’s ethical guidelines, the study involved 1,006 participants, primarily students, who reported high levels of loneliness and varying degrees of perceived social support. Notably, 90% of participants experienced loneliness, a significant increase compared to previous studies, suggesting that Replika may serve as a crucial resource for mental health support among this demographic, particularly given their limited access to traditional counseling services.
The outcomes of using Replika were categorized into four types, with many participants reporting positive effects, such as reduced anxiety and, for some, even prevention of suicidal ideation. The Selected Group, comprising individuals who credited Replika with preventing suicide, exhibited a strong correlation between feelings of loneliness and perceived social support. This group also demonstrated a greater likelihood of viewing Replika as a human-like intelligence, which may enhance their engagement and the perceived efficacy of the application. The findings indicate that ISAs like Replika can stimulate human relationships rather than displace them, suggesting their potential as supportive tools in mental health contexts. However, the study also calls for further investigation into the long-term impacts of ISA interactions on mental health, particularly concerning ethical considerations and the risks associated with their use.
