Search for: All records

Online peer-to-peer therapy sessions can be effective in improving people's mental well-being. However, online volunteer counselors may lack the expertise and necessary training to provide high-quality sessions, and these low-quality sessions may negatively impact volunteers' motivations as well as clients' well-being. This paper uses interviews with 20 senior online volunteer counselors to examine how they addressed challenges and acquired skills when volunteering in a large, mental-health support community - 7Cups.com. Although volunteers in this community received some training based on principles of active listening and motivational interviewing, results indicate that the training was insufficient and that volunteer counselors had to independently develop strategies to deal with specific challenges that they encountered in their volunteer work. Their strategies, however, might deviate from standard practice since they generally lacked systematic feedback from mentors or clients and, instead, relied on their personal experiences. Additionally, volunteer counselors reported having difficulty maintaining their professional boundaries with the clients. Even though training and support resources were available, they were underutilized. The results of this study have uncovered new design spaces for HCI practitioners and researchers, including social computing and artificial intelligence approaches that may provide better support to volunteer counselors in online mental health communities. 

Online mental health communities (OMHCs) have emerged in recent years as an effective and accessible way to obtain peer support, filling crucial gaps of traditional mental health resources. However, the mechanisms for users to find relationships that fulfill their needs and capabilities in these communities are highly underdeveloped. Using a mixed-methods approach of user interviews and behavioral log analysis on 7Cups.com, we explore central challenges in finding adequate peer relationships in online support platforms and how algorithmic matching can alleviate many of these issues. We measure the impact of using qualities like gender and age in purposeful matching to improve member experiences, with especially salient results for users belonging to vulnerable populations. Lastly, we note key considerations for designing matching systems in the online mental health context, such as the necessity for better moderation to avoid potential harassment behaviors exacerbated by algorithmic matching. Our findings yield key insights into current user experiences in OMHCs as well as design implications for building matching systems in the future for OMHCs. 

Many AI system designers grapple with how best to collect human input for different types of training data. Online crowds provide a cheap on-demand source of intelligence, but they often lack the expertise required in many domains. Experts offer tacit knowledge and more nuanced input, but they are harder to recruit. To explore this trade off, we compared novices and experts in terms of performance and perceptions on human intelligence tasks in the context of designing a text-based conversational agent. We developed a preliminary chatbot that simulates conversations with someone seeking mental health advice to help educate volunteer listeners at 7cups.com. We then recruited experienced listeners (domain experts) and MTurk novice workers (crowd workers) to conduct tasks to improve the chatbot with different levels of complexity. Novice crowds perform comparably to experts on tasks that only require natural language understanding, such as correcting how the system classifies a user statement. For more generative tasks, like creating new lines of chatbot dialogue, the experts demonstrated higher quality, novelty, and emotion. We also uncovered a motivational gap: crowd workers enjoyed the interactive tasks, while experts found the work to be tedious and repetitive. We offer design considerations for allocating crowd workers and experts on input tasks for AI systems, and for better motivating experts to participate in low-level data work for AI.