Mobile CBT-I Sleep App Prototype

ISEF Category: Translational Medical Science

Subcategory: Disease Treatment and Therapies  ·  Difficulty: Intermediate  ·  Setup: Home Setup  ·  Time: 1 to 2 Months

The Hook

Sleep apps promise a lot, but most never prove they help. You can test whether a tiny CBT-I style intervention actually changes sleep data from a wearable. That turns a phone idea into a real experiment. It also teaches you how behavior, data, and health connect.

What Is It?

CBT-I means cognitive-behavioral therapy for insomnia. It helps people change habits and thoughts that keep sleep broken, like irregular bedtimes, worry about sleep, and too much time awake in bed. A micro-intervention is a very small version of that support, usually built into a phone app with one or two prompts, reminders, or exercises.

Think of sleep like a race car pit stop. If the stop is sloppy, the car loses time for the whole race. CBT-I tries to fix the pit stop, not just the car. Your project asks whether a short daily nudge, like a wind-down prompt, a consistent bedtime reminder, or a sleep diary check-in, changes sleep outcomes that a wearable can measure.

Wearables such as Apple Watch or Fitbit estimate sleep stages, total sleep time, and wake after sleep onset. Those numbers are not perfect, but they can still give you a consistent way to compare nights. Your job is to design the intervention, track the data, and test whether the pattern changes over 14 days.

Why This Is a Good Topic

This is a strong science fair topic because you can test a clear behavior change with measurable outcomes. You do not need a hospital or animal model. You can build a simple intervention, collect wearable sleep data, and analyze patterns across days or people. The project connects to a real problem, insomnia, that affects school performance, mood, and health.

Research Questions

• How does a nightly CBT-I style reminder affect total sleep time over 14 days?
• What is the effect of a consistent wind-down prompt on sleep onset latency?
• Does a sleep diary check-in improve sleep efficiency compared with no check-in?
• To what extent do bedtime reminders reduce night-to-night variability in wake time?
• Which micro-intervention, a reminder, a diary prompt, or a relaxation cue, produces the largest change in wearable sleep-stage metrics?
• How does the timing of the intervention affect next-night sleep quality?

Basic Materials

• Smartphone with app-building or survey capability.
• Apple Watch or Fitbit with exported sleep data.
• Spreadsheet software for tracking nightly outcomes.
• Sleep diary template.
• Participant consent form.
• Alarm or reminder app.
• Stable charging cable for the wearable.
• Notebook for observation logs.

Advanced Materials

• Wearable device with raw or exportable sleep-stage data.
• Phone app prototyping tool.
• Secure cloud folder for de-identified data storage.
• Statistical software for repeated-measures analysis.
• Python or R for data cleaning and visualization.
• ImageJ only if you need to inspect exported charts or screenshots.
• Validated insomnia questionnaire for pre and post comparison.
• Codebook for intervention logging and outcome labeling.

Software & Tools

• Google Sheets: Organizes nightly sleep metrics and helps you graph changes over time.
• Python: Cleans wearable exports, aligns dates, and runs repeated-measures analysis.
• R: Runs simple statistical tests and makes clear plots for small sleep datasets.
• JASP: Lets you compare conditions with point-and-click statistics.
• Qualtrics: Collects brief daily sleep check-ins and intervention responses.

Experiment Steps

1. Define the single sleep outcome you will treat as your main endpoint, such as sleep onset latency or sleep efficiency.
2. Choose one micro-intervention to test first, and keep the message or action consistent across the study.
3. Plan your baseline period so you can compare normal sleep against your intervention nights.
4. Build your data structure before you start, including dates, wearable metrics, self-reports, and notes about missed nights.
5. Decide how you will separate device noise from real change, such as by using repeated measures or within-person comparisons.
6. Preplan the control conditions, visualizations, and statistical test you will use before you look at the results.

Common Pitfalls

• Mixing sleep metrics from different wearable brands, which makes the endpoints hard to compare.
• Changing the app prompt during the study, which turns one intervention into several different ones.
• Ignoring device data gaps or failed uploads, which can hide missing nights.
• Using only one self-report question, which leaves you without a clear comparison to the wearable data.
• Comparing nights with very different schedules, which confounds the effect of the intervention with school nights, weekends, and travel.

What Makes This Competitive

A stronger project does more than ask whether a sleep app works. You can compare two or three micro-interventions, test the same idea in more than one person, or pair wearable data with a validated insomnia survey. Clean data handling matters a lot here, because sleep exports can be messy and noisy. Strong controls, clear endpoints, and careful statistics can turn a simple app test into a serious translational study.

Project Variations

• Test the same micro-intervention in student athletes to see whether training schedules change sleep response.
• Compare Apple Watch and Fitbit sleep-stage estimates for the same intervention nights to study device agreement.
• Add a mood or alertness survey the next morning to see whether sleep changes match daytime function.

Learn More

• NIH PubMed: Search review articles on CBT-I, insomnia, and wearable sleep measurement.
• NIH National Heart, Lung, and Blood Institute: Find patient-friendly and clinician-facing information on insomnia and sleep health.
• CDC Sleep and Sleep Disorders: Read basic public health guidance on sleep duration, sleep habits, and sleep loss.
• Apple Developer Documentation: Look for HealthKit sleep data documentation and export details.
• Fitbit Help Center: Find help articles on sleep stages, sleep logs, and data exports.
• MIT OpenCourseWare: Search psychology, data analysis, or digital health materials for study design ideas.