Health behavior research

Measure health behavior where it happens

Physical activity, eating, drinking, sleep, medication adherence — captured at the moment and in the place they occur, not reconstructed from memory a week later.

7

Scheduling modes

30

Cognitive tasks in-app

16

Participant-app languages

The measurement problem this solves

A weekly food-frequency or activity questionnaire asks participants to average over episodes they never encoded separately. What comes back is a reconstruction shaped by mood, salience, and how the question was worded — and the error is rarely random, because the people whose behavior is most variable are the ones least able to summarise it.

Momentary assessment replaces that summary with observation. The participant is asked while the behavior is happening, in the context that produced it, and the answer arrives with a timestamp and a location rather than a recalled week.

Designs you can run

Each of these is assembled in the study builder — no app development, no separate sensor toolchain.

Place-based consumption and adherence diaries

Place-visit detection fires when a participant arrives somewhere new, so you do not have to pin every supermarket, café, gym, or clinic in advance. Use it for eating occasions, drinking contexts, or attendance at a treatment site.

Place visitNo pre-registered locations

Geofenced just-in-time prompts

Define a location and trigger a survey or a message on entry or exit — a gym for an activity check, a workplace for a stress measure, a clinic for an adherence item. The prompt reaches the participant while the behavior is still in progress.

GeofenceEnter / exit

Answer-contingent follow-up

A survey-answer trigger fires a second survey when a participant gives a specific response — reports a lapse, a craving, a missed dose, a symptom flare. The follow-up arrives during the episode instead of at the next scheduled beep.

Event-basedSurvey answer trigger

Passive health data next to self-report

Collect steps, sleep stages, active energy, and walking and running distance from Apple HealthKit and Android Health Connect — plus resting heart rate, heart-rate variability, blood oxygen, and respiratory rate where the participant pairs a wearable — and read them against the momentary reports from the same day.

HealthKitHealth ConnectWearable-dependent metrics labelled

Sleep- and wake-anchored sampling

Anchor prompts to each participant's own wake time, sleep time, or preferred days and times, taken from prescreening, instead of a single clock that fits nobody. Useful whenever the behavior of interest follows the participant's day rather than yours.

Participant-anchoredPrescreening-based

Multi-week intervention and follow-up protocols

Long protocols live or die on adherence. Watch per-participant response rates by day, week, and survey type, set a threshold, and receive an email digest when someone falls below it — early enough to intervene rather than after the data is gone.

Compliance dashboardAdherence alerts

The parts reviewers ask about, built in

Design decisions that are hard to defend after the fact are easier to make before fielding.

Power planning for the design you actually built

The study builder links into the ESM Power & Design Planner with your participants, days, and prompts per day carried across. For event- and location-triggered schedules it does not invent a prompt rate, because that number is an outcome of the participant's week rather than a design choice.

Open the power planner →

A burden estimate before you field it

Time per prompt, prompts per day, and total minutes across the study are estimated from the design, checked against published benchmarks, and testable against alternatives without touching the live study. It exports a written burden statement and a participant disclosure for the ethics application.

Validated items rather than improvised ones

Build surveys from a library of validated scales and question modules tagged by construct, or add cognitive tasks with mobile-tuned trial counts and scoring. Every item is browsable publicly before you sign up.

Protocol out first, data out after

Export the protocol — surveys, items, schedules, estimated burden — as a machine-readable file and a printable document to attach to a preregistration. Export the data as CSV, JSON, or SPSS, in a long format that goes straight into a multilevel model.

Start from a health behavior template

Each template is a complete study configuration — surveys, items, and schedules — that you can preview publicly and clone into your account.

Browse all templates →
Health

Physical Activity EMA with Accelerometry

Interval-triggered EMA capturing physical activity context and perceived exertion alongside concurrent accelerometer and pedometer data. Background step count and heart rate via HealthKit. Suited for exercise science and behaviour change research.

1 survey · preview free

Health

Eating Behavior & Food Diary

Interval-triggered per-episode diary capturing hunger, fullness, eating rate, emotional state, and mindless eating. Suited to nutritional psychology and eating behaviour research.

1 survey · preview free

Addiction

Alcohol Use EMA

Daily evening alcohol diary with AUDIT-C adapted items and a weekend morning hangover check. Suited to public health, addiction, and epidemiological research.

2 surveys · preview free

Health

Stress & Physical Activity Monitoring

Morning exercise intention and stress prediction log, and evening perceived stress and physical activity review. Suited to health psychology and exercise science research.

2 surveys · preview free

Sleep Science

Sleep & Wearable Health Monitoring

Passive collection of sleep stages, resting heart rate and daily steps from Apple Health or Health Connect, combined with brief morning and evening self-report diaries. Designed for sleep science and digital health research.

2 surveys · preview free

Health

Cardiac Rehabilitation Monitoring

Daily symptom self-monitoring and weekly exercise progress review for cardiac rehabilitation patients. Suitable for cardiology and health behaviour research.

2 surveys · preview free

EU-hosted and GDPR-native

Data stays on EU infrastructure. Personal identifiers, consent records, and research data are kept separate, and a DPA is available for your institution.

Security and privacy

Encryption with a key you hold

Turn on per-study public/private key encryption and responses are encrypted on the device. Without the private key your study team holds, the stored data cannot be read.

How encryption works

Free tier, academic pricing, free for theses

A permanent free tier for piloting, 30% off every paid plan with an academic email, and a full sponsored plan for Bachelor's and Master's thesis projects.

See pricing

Questions researchers ask first

Can I collect data from a wearable?
SMAAT reads from the phone's health store — Apple HealthKit on iOS, Health Connect on Android — rather than integrating with each wearable directly. If a participant's watch or band writes to that store, SMAAT can collect it. Metrics that in practice need a paired wearable (resting heart rate, HRV, blood oxygen, respiratory rate) are labelled as such in the study builder so you do not design around data your sample will not produce.
Does the app keep collecting when it is closed or offline?
Yes. Passive health metrics sync on a background task, and survey responses collected without connectivity are stored on the device and uploaded once the participant is back online. Nothing is lost to a train tunnel.
Health data is a special category under GDPR. How is that handled?
Hosting is in the EU, personal identifiers and consent records are held separately from research data, and per-study end-to-end encryption is available with the private key held by your team. A data processing agreement is available for your institution, and the study builder produces a participant-facing disclosure you can attach to your ethics application.
How do I justify participant burden to an ethics committee?
The dashboard estimates the burden of your design before you field it — time per prompt, prompts per day, total minutes across the study — flags designs that look heavy against published benchmarks, and lets you test alternatives without touching the study. It exports a written burden statement and a participant disclosure for the ethics application.
Can I run a just-in-time adaptive intervention?
You can build the triggering and delivery side of one: geofence entry, place visits, and answer-contingent follow-ups all fire in real time, and you can deliver a survey or a message in response. SMAAT does not ship a decision engine that selects the intervention for you — the tailoring logic is yours to specify.

Field your next study on SMAAT

Start free, keep the free tier for piloting, and pay academic rates when the study goes live. Bachelor's and Master's thesis projects are sponsored in full.