"Actually I Can Count My Blessings": User-Centered Design of an Application to Promote Gratitude Among Young Adults

Young adults face increasing stress, anxiety, and mental health challenges driven by academic/career pressures, social media, and life transitions. Prevalence of mental illness is high in ages 18–25, and access to care can be limited. Gratitude practice is associated with improved mental and physical health, relationships, and life satisfaction, yet young adults struggle to integrate it into daily life due to unfamiliarity, lack of guidance, and time constraints. Prior gratitude tools are commonly top-down, emphasizing clinical expertise over user contexts, potentially reducing fit with daily routines. The authors propose a user-centered design (UCD) approach to identify features that help young adults incorporate gratitude into daily routines via a mobile app (chosen for ubiquity and capabilities like notifications). Research questions: (RQ1) What key features of a mobile app could promote regular gratitude practice among young adults? (RQ2) What are perceived benefits of using such apps, and how do benefits vary by engagement levels? The work includes a formative study using third-party apps to elicit needs and a deployment study of a custom app to test structured prompts, timing, and mood labeling.

The paper reviews three areas. (1) Positive impact of gratitude on wellbeing: Gratitude is an emotional and social experience of recognizing positive aspects of life and sources of goodness. Regular gratitude is linked to lower anxiety, depression, and loneliness; better coping and resilience; savoring positive experiences; and stronger social bonds through enhanced closeness, optimism, and kindness. (2) Interventions to promote gratitude: Offline methods (journals, letters, acts of kindness) improve life satisfaction and reduce stress and depressive symptoms. Online tools include mobile apps with prompts, reminders, multimedia entries, and sometimes social features; web journals; and social media campaigns/groups. AR/VR show potential but have lower ubiquity. Many tools are limited by top-down design, insufficient customization, and limited fit with users’ contexts. (3) UCD of digital mental health interventions: UCD gathers user needs via interviews, surveys, observations, and iterative testing. It is increasingly used to tailor interventions for young adults, supporting varied engagement and contexts (e.g., automated messaging tools, peer support apps, medication reminders). There is limited user-perspective research focused specifically on gratitude, motivating a UCD approach to identify critical features and design elements for gratitude practice.

Research overview: Two connected studies. Formative Study to elicit desired features using existing apps, followed by a Deployment Study of a custom-built gratitude app informed by formative findings.

Formative Study:

• Participants: 20 young adults (10 women, 10 men; mean age 21.2 ± 0.78) from a large European and a large North American city; inclusion: ≥18 years, smartphone owner.
• Apps used: Delightful – Three Good Things (Android) and Gratitude 365 (iOS). Selection criteria: simplicity, gender neutrality, high ratings, free availability, comparable design (no suitable cross-platform single app was found). Features included multiple daily text entries, viewing past entries/dates, reminders; minimal guidance; Delightful sent motivational quotes.
• Procedure: Participants used an assigned third-party gratitude app self-directed for two weeks; no usage tracking. Post-period semi-structured Zoom interviews (15–35 minutes) probed facilitators/barriers, helpful features, routine formation, and social contexts. Compensation: $10 USD. Ethics approvals obtained at two universities; interviewers trained in Columbia Suicide Risk Assessment; participants could skip questions or discontinue.
• Data analysis: Thematic analysis with two coders using open coding; iterative codebook development and consensus; final codebook applied across transcripts.

Key formative insights informing design: need for structured guidance/examples; evening/after-work timing and private contexts; benefits of mood labeling pre/post activity.

Custom Application Design (for Deployment Study):

• Platforms: Android and iOS.
• Two conditions: Random assignment to Experiment vs. Control. • Experiment: Life area options (derived from PERMA; e.g., Physical Wellbeing, Peace & Calm, Energizing Moments, Engagement/Flow, Connection, Accomplishment, Meaning/Fulfillment, Other) selected before writing entry; “More Info” link with explanations/examples (e.g., Connection). Then open text entry. • Control: Open text entry without life area selection.
• Both conditions: Mood labeling pre- and post-entry (single-item 1–5 scale). Daily reminder notification at 6 pm (fixed time to support habit formation); dark theme for privacy. App pages: Dashboard (past entries), Profile (stats), About (benefits and instructions), Settings (view reminder time, logout).

Deployment Study:

• Participants: 26 young adults (16 women, 9 men, 1 non-binary/third gender; mean age 22.3 ± 0.98), recruited similarly to formative study and from different individuals; 13 per condition. Randomized assignment blinded to participants.
• Procedure: Guided app installation support; two-week use; concluding survey (1–7 Likert) assessing perceived motivation, willingness to engage when not in a good mood, and perceived usefulness; post-study Zoom interviews (20–45 minutes) about notifications, mood change, life area options (experiment only), and feature ideas. Compensation: $10 USD.
• Quantitative data: Descriptive statistics on survey measures; app-derived engagement metrics including word count per entry and mood change per entry; retention (days with entries). Note: Passive engagement (reflection without entry) not captured by usage metrics.
• Qualitative data: Thematic analysis as in formative study.
• Ethics: App not a crisis service; crisis resources provided; daily monitoring of entries for self-harm risk with protocol for outreach using the Columbia Suicide Risk Assessment; no incidents occurred.

Formative Study themes:

• Desire for structure: Many participants (especially novices) struggled with open-ended journaling and wanted examples, explanations, and prompts to guide reflections; even experienced journalers valued prompts to avoid repetition.
• Timing and context: Preferred private, reflective contexts, most often evenings/before bed; routine and consistent cues encouraged adherence; dark/incognito theme desired for privacy.
• Mood labeling benefits: Pre/post mood reflection increased awareness, revealed mood shifts toward positivity, and could help identify triggers; participants endorsed brief, lightweight mood rating.

Deployment Study quantitative results (Experiment vs. Control):

• Word count per entry: 15.27 ± 2.42 vs. 8.90 ± 0.83 (higher in experiment).
• Mood increase per entry (post minus pre, 1–5 scale): 0.64 ± 0.04 vs. 0.56 ± 0.05 (higher in experiment).
• Perceived motivation (1–7): 4.46 ± 0.40 vs. 4.38 ± 0.40.
• Perceived engagement when not in good mood (1–7): 4.62 ± 0.46 vs. 4.08 ± 0.44 (higher in experiment).
• Perceived usefulness (1–7): 4.92 ± 0.24 vs. 4.92 ± 0.35 (equal).
• Retention: Entries on 79.1% of study days (experiment) vs. 76.4% (control); active participants comparable across days. Note: Metrics do not capture passive engagement; thus, results are suggestive rather than conclusive.

Deployment Study qualitative findings:

• Structured life areas: Experiment participants reported prompts helped focus and broaden reflections without constraining expression; balance of structured cues with open-ended text was valued. Several control participants desired more guidance.
• Notification timing: 6 pm prompts were appreciated and anticipated; often served as reminders to complete later in the evening; participants suggested integrating with personal routines (e.g., commute) or phone Focus modes.
• Mood labeling: Mixed responses—many valued awareness and observable mood improvement; some felt burdened labeling mood before and after each entry; suggestions included reducing frequency, making it optional, or delaying post-assessment (e.g., 1 hour later).
• Perceived benefits: Positive shifts in perception, increased appreciation, and potential resilience; habit spillover into daily life and social interactions; recognition of benefits from passive engagement (mentally reflecting without writing), though it leaves no record.

Design recommendations derived:

• Balance mood labeling frequency/timing and allow personalization.
• Support varied engagement including passive options and low-effort interactions; consider multimedia and careful incentives (e.g., streaks) without pressure.
• Encourage transfer to real-life contexts and social/communal practices (sharing, group boards).
• Personalize prompt timing using user schedules, calendars, and sensors (privacy-respecting) and allow user-set times.
• Scaffold for novices and offer advanced options for experienced users (custom prompts, goals).

Findings address RQ1 by identifying key features that promote regular gratitude practice among young adults: structured prompts using life area options grounded in PERMA; consistent, context-appropriate reminders (late afternoon/evening) to support habit formation; and lightweight mood labeling to enhance self-awareness. The structure reduced ambiguity, increased depth/diversity of reflections, and correlated with higher engagement (word count) and mood improvement per entry. Fixed-time notifications supported routine yet also acted as flexible reminders for later completion, suggesting benefits of both consistency and context alignment.

For RQ2, perceived benefits included reframing outlooks toward positivity, appreciating small moments, and potential resilience gains. Engagement varied from active (writing) to passive (thinking), with participants reporting benefits even from passive contemplation prompted by notifications. Habit formation extended beyond the app to interpersonal contexts (e.g., conversations with partners and friends), suggesting social ripple effects of gratitude practice.

The significance lies in demonstrating that UCD-driven features resonate with user contexts, increasing acceptability and perceived value. Mixed reactions to mood labeling underscore the need for personalization to avoid burden. Supporting passive engagement may mitigate attrition during busy or low-energy periods, broadening inclusivity. Overall, the study shows that aligning digital gratitude tools with users’ routines, preferences, and social environments can enhance engagement and wellbeing outcomes, informing HCI/CSCW design of digital mental health tools.

The study applied a UCD approach to design and evaluate a mobile application for gratitude practice among young adults. A formative study highlighted needs for structured guidance, evening prompts, privacy, and mood labeling. A two-week deployment of a custom app confirmed that structured life area options can increase engagement (higher word counts) and mood improvements per entry, and that consistent evening reminders support routine. Participants reported positive perception shifts, resilience, and habit spillover into daily and social life; passive engagement also offered benefits. The work offers concrete design considerations: personalize mood labeling frequency and timing, support varied engagement (including passive), facilitate transfer to real-life and communal contexts, adapt prompt timing to individual schedules and contexts, and tailor scaffolding to users’ experience levels. Future research should expand personalization, incorporate objective measures, explore broader demographics and cultural contexts, and examine additional design modalities (e.g., conversational guidance).

• Sample and context: Participants were drawn from one European and one North American city; cultural/contextual differences may limit generalizability. Focus on young adults (18–25) may not extend to other demographics (teens, middle-aged adults, immigrants) with distinct needs.
• Measures: Heavy reliance on self-reports introduces potential biases (response, social desirability). App metrics did not capture passive engagement; objective measures (e.g., physiological or behavioral) were not included.
• Design scope: The custom app explored a subset of possible features (e.g., no conversational/chat guidance, limited personalization of reminders/content, minimal multimedia). Advanced personalization (e.g., adapting to prior gratitude experience, schedule-aware prompts) was not implemented.
• Timing data: Fixed 6 pm notifications may not be optimal for all; no adaptive timing was tested.
• Duration and power: Two-week deployment and modest sample size limit detection of longer-term effects and statistical inference; reported quantitative differences are descriptive/suggestive.