Dwelling: A Wellness App

Climbing out of catastrophisation spirals, one step at a time

During the summer, I came up with a case study inspired by my experiences to create a concept for an app that helps students to stop unnecessary rumination on past events and worrying about future events (Ruminative Negative Thinking - RNT).

Team

Just me for this one!

Tools

Figma, Miro, Canva

Role

End to end product designer &user researcher.

Skills

User research, concept ideation, interaction design, prototyping, user testing

THE PROBLEM

How can we make students motivated to change their behaviour while catastrophising?

Undergraduate students frequently catastrophise.

University is a high-risk period for stress (Abrams, 2022). Many students experience catastrophisation (spiralling), as exemplified by the current 'crashing out' trend on Tiktok, which leads to irratioinal decision making and lower levels of mental health.

Distraction, mindfulness & CBT effectively stop catastrophisation.

Distraction, mindfulness and CBT can reduce unhelpful ruminative negative thinking (RNT patterns).

They are more effective when completed together than on their own and require persistent use to be fully effective (Shen et al.).

However, students rarely engage in these during catastrophisation.

Up to 21 - 54% of meditation app users drop out within the first few months of downloading, and the daily use of health apps among paying subscribers is less than 4% (Sullivan et al, 2023). Integrating app use into a daily routine contributes to more persistent use.

THE SOLUTION

An app that offers a linear 4 stages of help and gamifies mental wellbeing as a currency.

Click here to see interactive prototype! 👀

A guided 4 step process to empower users when feeling down.

Equipping users with a predetermined plan (relax + rethink + react + recognise) so that they can take action against their RNT with minimal effort when they need it most.

1. Gamified mindfulness activities to help users regain control.

Helping users regain control by grounding their minds with a quick mindfulness exercise of their choice, reducing emotional overwhelm.

2. A psychology backed reflection task to reframe negative thoughts.

Guiding users to reframe unhelpful thoughts, making space for more balanced and constructive perspectives based on reflections of their own situation.

3. A timed segment pushing users to get off their phone.

Calling users to action with an interactive timed activity, breaking their loop of negative thinking through intentional redirection of thoughts.

4. Interactive learning modules to be completed when feeling better.

Empowering users to easily and accessibly learn about RNT, building long-term resilience and self-awareness for future rumination or worrying.

How did I get there?
Let's find out...

1. Discovering
the problem

1.1 Research
questions

SECONDARY RESEARCH

A few years ago, I visited my doctor, overwhelmed by constant worry (classic second year uni experience!). His response was straightforward—he explained that I was experiencing unhelpful thought patterns and handed me a link to a website called This Way Up, which offered structured cognitive behavioral therapy (CBT) modules. That was it. No further guidance, no follow-up—just a resource and a hope that I’d stick with it.

In fairness,  gave the program a try, but the sessions were long and time-consuming. While I recognized that the content was valuable, I quickly lost motivation and stopped using it. The problem wasn’t the information itself—it was the way it was delivered.

Over time, I did find my own ways to cope with my anxieties (doing hobbies helped, and talking to my friends even when I didn't feel like it. And lots of midnight internet search sessions about how cognition works). But this experience led me to a key question: Could digital tools have supported me more effectively? I set out to explore this through secondary research, conducting a literature review and competitive analysis to understand existing solutions and their limitations. Here’s what I found:‍
‍

PRIMARY RESEARCH

While my experience definitely aligned with my research, I wanted to hear of my target audience's experiences firsthand to ensure I wasn't a minority in this issue and to cater the final product toward their needs. To do so, I used the triangulation method of study, engaging in questionaires, semi structured interviews and online ethnography.

60 responses | Age 18 - 25

20 question Questionaire on Google Forms (Short responses and multiple choice)

8 responses | Age 18 - 24

Semi - structured Interviews on Zoom (With 8 participants, 15 minutes each)

30+ Online forums

Online Ethnography on various forums (Reddit, Quora and mental health forums)

1.2 Research
findings

KEY INSIGHTS

To make sense of my research findings, I organized them into an affinity diagram, grouping similar themes and patterns. As I sorted through user frustrations, behavioral tendencies, and current student responses to catastrophisation, clear categories began to emerge: sers struggled with overwhelming content, rigid structures, and a lack of real-time support when they needed it most, but were generally to try new i. This process helped me identify the core pain points that tot address—designing an intervention that was intuitive, quick, and emotionally supportive rather than another task on a to-do list.

Below is a synthesised version of the affinity diagram after grouping common bottom level insights into common points (there were originally 100+ bottom level insights that I summarised into a few key points here).

I established 4 main themes:

1. Causes of RNT

RNT was typically triggered by concerns around exams, work and relationships - and were more common around stressful periods like exam or internship season.

2. Motivation to change behaviour

My interviewees had trouble being motivated to change their attitudes or behaviour whilst in the RNT cycle as they did not realise they were engaging in RNT.

3. Attitude to CBT and learning

My interviewees were generally very open to meditation and CBT as interventions. Most participants did not understand what RNT was and expressed a desire to learn more, but only if it was easy to understand.

4.  Performance of Solutions

Participants were more likely to engage in interventions when they recognised their thoughts as unhelpful and when there was external pressure - i.e. when they had a prior commitment.

So in summary - students were open to mobile methods of targeting their RNT, but found it difficult to remain open to doing so when actually ruminating/ worrying, which mostly happened during stressful periods. The most effective experiences of snapping out of a RNT cycle were when they had extrinsic motivation to stop ruminating/ worrying. These insights led me to refine my problem statement, which would inform the rest of my processes:

Refined Problem Statement

How can I make university students feel more motivated to change their attitudes while engaging in Repetitive Negative Thinking (RNT)?

Learnings: Discovery phase

You've reached a reflection checkpoint!

Rest your eyes while you read about some key learnings I made along the way - for next time 🤔💭

My primary research was too broad 🥸

One of the key issues was that my primary research was too broad. While I aimed to understand catastrophization from multiple perspectives, the wide scope made it difficult to extract precise, actionable insights.

In hindsight, a more targeted approach—such as focusing on specific triggers for catastrophization or narrowing down user demographics—could have yielded clearer patterns and stronger design decisions.

My interview participants were too limited 👪

Although the survey and interview responses provided valuable qualitative insights, a larger and more diverse participant pool would have helped validate the findings more robustly.

Expanding the research to include individuals with varying anxiety levels, professional backgrounds, and coping strategies could have provided a more comprehensive understanding of user needs.

I didn't account for personal bias 🪞

As someone deeply invested in mental health, I found myself unconsciously gravitating toward solutions that aligned with my preexisting beliefs.

While I made efforts to remain objective, incorporating more external validation—such as collaborating with mental health professionals— could have mitigated this bias. Going forward, I will prioritize a more structured and systematic approach to data collection and analysis.

2. Concept
Ideation

2.1 Ideation methods

STORYBOARDING & DECISION MATRIXES

To generate a diverse range of potential solutions for my project, I first engaged in Reverse Thinking. I asked myself what ideas would be the most 'cringe' to students in a bad mood, what measures would not be encouraging to use, and what solutions would make someone who is catastrophising feel worse.

I then engaged in Crazy 8s to brainstorm eight different concepts in eight minutes, focusing on ways to help users escape catastrophization cycles. The goal was to explore unconventional ideas without overanalyzing feasibility too early. After reviewing the ideas alongside my Reverse Thinking, I narrowed them down to three promising directions, below.
‍
I then evaluated each idea using a decision matrix, scoring them across multiple criteria to determine the most practical and impactful solution. This exercise highlighted potential challenges—such as accessibility issues, high production costs, and engagement concerns—which helped me improve and refine each concept before making a final decision:

OPTION 1

A four in one gamified wellness app

A digital experience that turns cognitive reframing into an engaging, interactive process. Instead of traditional journaling or therapy-style modules, users navigate challenges, unlock insights, and reframe their anxious thoughts through guided, bite-sized tasks.

OPTION 2

A VR meditation and exercise room

A virtual space where users can physically "walk through" their thoughts, combining mindfulness with light movement exercises. This immersive experience could help users externalize their anxieties and gain perspective in a fully controlled, distraction-free environment.

OPTION 3

A wellness plushie that sends friendly reminders

A tangible, interactive plush companion that responds to user input. It could guide breathing exercises, provide gentle vibrations for grounding, and offer simple, comforting phrases to help disrupt anxious spirals

3. Defining
the problem

3.1 Who am I
designing for?

USER ARCHETYPES

Now that I'd clearly defined the issue and a possible solution, I wanted to ensure my possible end solution would reflect the needs of those who would actually use it. To this end, I analysed the experiences and demographic data of my interviewees and translated them into three personas characterised by how succeptible they are to RNT.

Anxious Ariana

Sometimes I get myself down, like I'll just think about something too much and it becomes such a problem when it's not, you know, not really that deep.

2nd Year • BArts • 19

Skeptical Selina

I don't know. All the meditation stuff makes sense and seems obvious in theory but then doesn't really make sense when you try it yourself.

4th Year • LLB • 21

Busy Barnie

I get through stressful periods ok, but not... effectively? I just have to kind of push through my stress and it's not fun but I always make it out alive somehow.

3rd Year • BEng  â€˘ 22

JOURNEY MAPS

Borne from my personas, I created journey maps. In these, I could better understand their goals, motivations, and frustrations. This holistic perspective ensured that every decision I made was aligned with user needs. Here is one for 'Anxious Ariana':

By understanding the pain points and behaviors of these archetypes, I determined that my app would have to be:
‍
- Fast & Actionable – Short, structured prompts that fit into daily life and are interactive.
- Research-Backed & Practical – No fluff, just clear sources and phrasing.
- Non-Intimidating – A friendly, supportive tool rather than a rigid program.

Creating these personas helped ensure that the app addresses real user needs, making cognitive reframing feel easy to integrate into everyday life (as the main response to my problem question). These three principles governed my design process, and I constantly returned back to them when designing my wireframes, lo - fi and hi - fi prototypes.
‍

📌 Problem statement - personas and journey maps

You've reached a reflection checkpoint!

Rest your eyes while I reflect on some improvements to note for next time:

I could have ideated alongside feedback

Next time I would test my concepts early on with real users while coming up with them.

I conducted my ideation and prototyping at two very different stages of the process. However, instead of only relying on theoretical feedback, I could have conducted low-fidelity prototype tests or focus groups to gauge initial reactions to the ideas themselves.

Feedback loops could help me quickly discard ideas that don’t resonate or refine concepts before fully committing to one direction.

I could have created empathy maps

In addition to traditional persona creation, empathy maps could help me fill in gaps in understanding.

Due to time pressure I did not have time to map out my personas' thoughts, feelings, actions, and pain points in a more visual and holistic way. For instance, under “Feelings” for Ariana, the empathy map might reveal that despite using the app, she still feels a lingering sense of uncertainty about her future or the outcome of her efforts- understanding these silent emotions would help me better tailor the experience for her in a more realistic way.

I could have made contrasting journey maps

A final key improvement I’d make is to create two separate journey maps for each persona: one for their experience before using my app and one for their experience after using the app.

This would allow me to more clearly compare the emotional shifts and pain points users encounter before they start using the app, and the transformation they undergo once they’ve integrated it into their daily routines.ting the personas based on real-world feedback would ensure that the app continues to serve their changing needs.

4. Designing
the solution

4.1 First Round of User Testing

EARLY LO-FI MOCKUPS & FIRST ROUND ITERATIONS

From the start, I wanted to engage in vigorous user testing to ensure I captured their needs and produced the most streamlined app possible. I created some lofi mockups and conducted research using:

  • Think aloud
  • Interviews (pre and post test)
  • Observations
  • Cognitive walkthrough

Based on the results, I made several iterations to address user feedback. The response to my wireframes and lo-fi mockups alerted me to several major issues embedded in the very structure of the app, so I made the following major changes:


4.3 Hi fi Prototypes and User Testing - Round 2

HEURISTIC TESTING

After making those changes, I created some mid fidelity mockups that I then tested with 4 experts (UX/UI design students) using a heuristic evaluation template.

This was to identify violations to usability standards as they allocated a severity rating & recommendation to identified usability issues. I identified that my top violated heuristics which needed to be iterated upon most were:

Violated Heuristic 1

Visibility of system status

‍Possible Violation: Users may not always know where they are in the 4-step process or if their progress is saved.

‍Fix: Use a clear progress indicator and visual cues to show completed steps. Also, provide immediate feedback after actions.

See how I implemented these changes on the 'Rethink' page iterations.

OPTION 3

Recognition rather than recall

Possible Violation: Users may have to navigate through multiple layers to access their desired option.

‍Fix: Present all relevant options clearly on the screen with intuitive navigation and group related items together.

See how I made menus clearer and features more visible in the 'Relax' page iterations below.

OPTION 3

Readibility and legibility

Possible Violation: Long paragraphs or hard-to-read fonts might strain the user’s eyes, especially in reflection or educational content.

‍Fix: Use concise, easy-to-read text with ample white space and break longer options into various pages.

See how I did this on both the 'Relax' and 'Rethink' pages below.

USER TESTING

I also conducted user testing with the table below, again conducting pre and post interview tests, observation, think aloud and cognitive walkthrough. I started by defining key tasks for the users to complete within the app and observed how they interacted with it. I created the below table to give myself a structured framework for evaluating both qualitative and quantitative data, making it easier to see which features were successful and which needed improvement.

5. Delivering
the final product.

Click here to see interactive prototype! 👀

ONBOARDING

Homepage

Stage 1: Relax

Stage 2: Rethink

Stage 3: React

Stage 4: Recognise

6. Reflection & Learnings

6.1 Success Metrics

Dwelling received an 88% (Excellent) SUS score from users, and a 4.8/5 on how likely anxious students would recommend it. The overall positive response from target users indicated that Dwelling would be an app students were willing to try out, and also use frequently to deal with their emotions more healthily on a situational basis. If Dwelling was to get developers involved, I'd consider these key success metrics:

Completion of Cycles

This would measure the percentage of users who successfully complete an entire cycle once starting it.

This will gauge how effectively users are engaging with the app’s core functionality, ensuring they are getting value from it.

I could implement a goal like 'Achieve 80% task completion rate once a cycle has begun.'

Customer Satisfaction

This metric would measure the percentage of users who rate their experience with the app as positive.

High satisfaction indicates the app is meeting user needs and has the potential for high retention.

I would set a success metric: 85% or higher CSAT score (measured through in-app surveys after completing key tasks).

Mental Health Improvement

We would benefit from measuring improvement in users' mental health or ability to cope with stress or anxiety over time, based on self-reported data.

This is the ultimate measure of Dwelling’s success.

We could aim for 70% of users report feeling less stressed or better able to manage catastrophizing thoughts after one month of use.

6.2 Things to note
for next time

This being a sprint passion project that I had to do between university assignments and the like, I felt quite rushed throughout the entire process. Next time I would definitely make some changes

Broader research demographics

While I conducted some user testing and initial research, I feel that the discovery phase could have been more thorough. In particular, there was limited engagement with a broader range of potential users and their needs in terms of mental health and repetitive negative thinking.

In future projects, I will engage with a wider variety of users based on the topic of my project.

Refining personas

While I did create personas to represent different user types, I felt that they were too generalized and didn’t fully capture the variety of needs or behaviors across users.

In the next project, I would refine personas with more granularity by including more behavioral data, motivations, and goals based more heavily on actual quotes and results from my surveys.

More structured iteration

In some cases, I created high-fidelity prototypes before testing the basic structure or functionality with users. This resulted in more effort upfront than was necessary, and some designs ended up needing significant changes after testing.

For future projects, I would focus on creating more detailed low-fidelity prototypes early in the process for quick, iterative user testing.

Thanks for reading!

Some other projects for you to check out...

1. Discovering
the problem

1.1 Research
questions

SECONDARY RESEARCH

Story time:
‍A few years ago I went to my doctor to ask what I could do about my constant worry about almost every area in my life. He essentially told me that what I was experiencing was unhelpful thought patterns, handed me off a link to a trial on a website called This Way Up, and sent me on my way. I tried the website for a few sessions, but ultimately the modules were long and time consuming and I lost all motivation to continue them. I just kind of stopped using it, even though I could tell the information was probably helpful 😳

Initial findings:
It's not like those problems went away, but over time I figured out my own ways to cope. Was there a better way apps could have been used to help me back then? I set out to engage in some secondary research, through a literature review and competitive analysis - this is what I found: ‍
‍

PRIMARY RESEARCH

While my experience definitely aligned with my research, I wanted to hear of my target audience's experiences firsthand to ensure I wasn't a minority in this issue and to cater the final product toward their needs. To do so, I used the triangulation method of study, engaging in questionaires, semi structured interviews and online ethnography.

Please press on the images below to see how I conducted my research! 👀

60 responses | Age 18 - 25

20 question Questionaire on Google Forms

8 responses | Age 18 - 24

Semi - structured Interviews on Zoom

30+ Online forums

Online Ethnography on various forums

1.2 Research
findings

KEY INSIGHTS

I sorted my insights into a bottom up affinity diagram, which helped me to identify trends arising from the interviews, survey and online forums. Below is a synthesised version of the affinity diagram after grouping common bottom level insights into common points (there were originally 100+ bottom level insights that I summarised into a few key points here).

Through my affinity diagram, I established 4 main themes:

1. Causes of RNT

RNT was typically triggered by concerns around exams, work and relationships - and were more common around stressful periods like exam or internship season.

2. Motivation to change behaviour

My interviewees had trouble being motivated to change their attitudes or behaviour whilst in the RNT cycle as they did not realise they were engaging in RNT.

3. Attitude to CBT and learning

My interviewees were generally very open to meditation and CBT as interventions. Most participants did not understand what RNT was and expressed a desire to learn more, but only if it was easy to understand.

4.  Performance of Solutions

Participants were more likely to engage in interventions when they recognised their thoughts as unhelpful and when there was external pressure - i.e. when they had a prior commitment.

So in summary - students were open to mobile methods of targeting their RNT, but found it difficult to remain open to doing so when actually ruminating/ worrying, which mostly happened during stressful periods. The most effective experiences of snapping out of a RNT cycle were when they had extrinsic motivation to stop ruminating/ worrying.

These insights led me to refine my problem statement, which would inform the rest of my processes:

Refined Problem Statement

How can I make university students feel more motivated to change their attitudes while engaging in Repetitive Negative Thinking (RNT)?

📌 Learnings - Discovery phase

You've reached a reflection checkpoint!

Rest your eyes while you read about some key learnings I made along the way - for next time.

Primary research was too broad

What are the main ways university students study for an upcoming exam?

1.1.2 Secondary Research (Method)

What are the key challenges students experience when trying to learn?

1.2 Secondary Research (Results)

What are the key challenges students experience when trying to learn?

2. Concept
Ideation

2.1 Ideation methods

STORYBOARDING & DECISION MATRIXES

Using crazy 8s to rapidly generate possibilities and then reverse thinking to refine these solutions, I narrowed down my solution to 3 possible ideas, which I then analysed using a decision matrix. I ended up choosing option 1 (interactive postbox) because of its resource efficiency, permanency and customisability for people of all social comfiness.

OPTION 1

A four in one gamified wellness app

This ranked highest on accessibility, fun, feasibility and interactiveness.

OPTION 2

A VR meditation and exercise room

While this idea was fun, it was not as accessible or feasible compared to other options.

OPTION 3

A wellness plushie that sends friendly reminders

This idea was the most feasible but seemed unhuman and was tone deaf to our issue.

Discovering the Problem.

Research questions

SECONDARY RESEARCH

Our secondary research involved 20+ papers, websites and news articles around the broad topic of study methods and rates of university course completion. We found, generally, that:

Time poor students perform worse in exams 😥

Time - poor students who do not attend lectures or spend time on self - study perform worse in assessments and exams

Method of study can predict exam success 🌱

Method of study may be more important than the time spent studying. Three useful study techniques include:

  • Generating own material (ie, summarisation, self explanation).
  • Visualisation (ie, flowcharts, mindmaps and diagrams)
  • Self testing (ie, practice papers, presenting to others).

PRIMARY RESEARCH

The secondary research left unclear, from a user perspective, how often these effective methods of studying were used, whether time- poor students benefited from them particularly, or why students may not use these.

Therefore, we used the triangulation method (surveys, semi- structured interviews and online ethnography) to develop our research question and give our project more direction.

💡 Our Research Question

To understand the methods and frustrations of students when trying to understand and digest large amounts of course content in preparation for exams and assessments.

We focused on the following sub questions to draw deeper insights:

BEHAVIOURAL

✍️ Current methods of studying

What are the main ways university students study for an upcoming exam? How do these methods affect their experience of exams?

BEHAVIOURAL & ATTITUDIONAL

😖 Challenges while studying

What are the key challenges students experience when trying to learn and revise new content?

ATTITUDIONAL

👀 Attitudes to diagramming

What are university students' perspective on using a mixture of diagrams and notes for studying?

Research findings

KEY INSIGHTS

Our research yielded some interesting results. After conducting our research, we used affinity diagramming to reach a few key insights: link to affinity diagram. We noted 3 key points from our research:

CURRENT METHODS OF STUDYING

Students mainly use note taking to study ✍️

'I study by taking notes from lectures and readings on Google Docs, Onenote or Notion' - Joanna, 21, Sydney

Students mainly study using note - taking, writing notes in their own words, copying slides and teaching others. To summarise information, they often use Google Docs, Notion and ().

CHALLENGES WHILE STUDYING

Students find volume of content overwhelming 😖

'I feel confused by overwhelming amounts of course content to learn in short periods' - Online ethnographic source

Students find understanding overwhelming amounts of content the most challenging part of studying. They have particular difficulties keeping up with lectures, revising and (). This leads to stress, a sense of being overwhelmed, and less desire to study.

ATTITUDES TO DIAGRAMMING

Students find visualisation helpful but impractical 🤔

'I find visualisation useful but making diagrams is hard to integrate into my study' - John, 21, Sydney

Many students find diagramming useful even though they don't directly use it in their studying. Students may not use diagramming in their studying because they feel it is too difficult to integrate with their existing notes, there are no good free diagramming software, or they feel that creating diagrams slows them down.

These insights led us to refine our problem statement:

💡 Refined Research Question

How do we integrate a hybrid solution to visual learning and traditional notetaking into students' study experience and encourage generation of own notes and self testing/ presentation?

Defining the Problem

Who are we designing for?

USER ARCHETYPES

We analysed the experiences and demographic data of our interviewees and translated them into three personas characterised by how much time and effort they dedicate to studying. Our target users lay in the range of not studying at all, to studying almost every day. Their reasons for not studying were mostly time poorness, ranging from social events, part time job

Rushed Ron

Studies ~2 hrs / week

I want to pass but because of my job I just have no time to study often.

2nd Year • BArts • 19

Studious Sue

Studies 6+ hrs / week

I try to get HDs but there is often too much content to study for at once!

4th Year • LLB • 21

Chill Guy

Studies <1 hr / week

Study isn't a top priority, I guess. I wouldn't stay home for it.

3rd Year • BEng  â€˘ 22

Competitor analysis

OVERVIEW OF COMPETITION

After narrowing down our concept, it was time to research competitors.

Docs

Note taking tool

3rd Year • BEng  â€˘ 22

Miro

Mindmapping tool

2nd Year • BArts • 19

Notion

Hybrid tool

4th Year • LLB • 21