DIGITAL WELLBEING
Research for digital wellbeing
Read original research on emotional support, self-understanding, and companionship. Each article distinguishes its question, method, findings, and limits from proposed applications to digital wellbeing.
Beyond app use: does mindfulness practice become part of daily life?
A US college trial measured app use separately from everyday practice and perceived skills. Orientation groups used the app more and reported some more favorable outcomes, but use, learning and changes in health need their own evidence.
Research commentaryGame-based interventions and implementationWhen a mental health game enters schools, delivery matters too
A randomized SPARX trial involving 134 Australian schools and 6388 students found no added reduction in depressive symptoms at 12 months. Twenty-two percent completed at least four modules; 12% completed all seven. Digital content needs evaluation alongside engagement, delivery conditions and the people it is intended to support.
Research commentaryImmersion and relaxationDoes feeling more immersed mean feeling more relaxed? A VR beach experiment
One hundred adults listened to the same breathing meditation with audio alone, a beach on a monitor, or a 360° VR beach. VR increased presence, but did not establish an added relaxation benefit. Immersive wellbeing experiences need separate tests of presentation, immediate experience and change over time.
Research commentaryInteractive narratives and experience evaluationHow can an interactive story respond to the person using it?
A four-week, 26-person Betwixt user-experience study offers clues about stories, pauses and responses. For Six Realms, relevance, participation and later experiences need separate evaluation; immersion and liking do not establish efficacy.
Product insightsEvaluating bedtime experiencesWhat can bedtime sound offer? From experience to outcome evaluation
In a four-week randomized trial of 495 adults, reported sleep disturbance improved in bedtime-audio and same-app daily-content groups, without clear additional benefit of bedtime audio on the primary outcome. Digital wellbeing evaluation needs a clear aim, comparison, and outcome.
Product insightImmediate and sustained supportHow long can digital companionship help? From a moment to daily life
A seven-day experiment observed lower momentary loneliness after chatting, but no sustained advantage over the comparison group before each day’s chat. Immediate support has value; digital wellbeing research should distinguish within-session, between-session, and post-use change.
Research insightTrust and understanding rolesTrust in digital wellbeing: evaluating closeness alongside reliability
Human-like expression can affect immediate trust and responsibility judgments without establishing greater reliability. Digital wellbeing research should consider perceived closeness, actual performance, and whether people understand the system’s role and accountability.
Research insightEmotional supportFeeling heard: what emotional support can an AI response offer?
Feeling understood is one starting point for digital wellbeing research. An experiment with 455 participants separated response content from source labels; the resulting immediate experience should be evaluated separately from lasting psychological change.
Research insightTrust and autonomyBetween empathy and agreement: leaving room to reflect
A comforting response can also reinforce a one-sided judgment. Research on AI sycophancy suggests evaluating emotional support alongside reflection, autonomy, and willingness to repair relationships.
Research insightCompanionship and connectionAI companionship in everyday life: support, relationships, and context
A survey of 1,131 users examined AI use alongside offline social context. Primarily using AI for companionship was associated with lower wellbeing; understanding that association requires attention to patterns of use and individual circumstances.
Research insightEvaluating sustained experienceLife after digital companionship: lessons from a four-week experiment
A four-week experiment with 981 participants detected no significant effects of assigned chat modes or topics. Greater voluntary use was associated with worse psychosocial outcomes, highlighting the need to distinguish experimental comparisons from behavioral associations.
Research insightStories and self-understandingAutonomy in interactive stories: what meaningful choices change
In a small game study, older adults with control over key story decisions reported greater autonomy and interest, alongside greater tension. Digital wellbeing research should distinguish these outcomes and examine whether changes continue beyond the experience.
Product insightResearch and evaluation methodsEvaluating emotional support: from satisfaction to sustained change
Evaluation should consider empathy, judgment, autonomy, and what follows a conversation. This article proposes a research approach informed by public studies, distinguishing methodological suggestions from established findings.
Product insightSee each article for evidence labels, samples, and sources. Our methods