Sensor-Triggered (Context-Aware) Experience Sampling

Description

Use phone, wearable or device-state data to decide when to prompt, so questionnaires arrive at moments of interest or at moments when the person can answer.

Choosing when to prompt balances coverage of daily life, enough instances of the target event, burden and compliance. Device-contingent sampling, introduced as context-aware ESM by Intille and colleagues, lets sensors or device state make that decision. Triggers can target events of interest (a smartwatch detecting a meal, an audio snippet from the Electronically Activated Recorder) or opportune moments (after a call or message, at activity transitions, never mid-sport or while driving). Opportune triggering has improved compliance and data quality and shortened response times. Whole-day schedules can follow inferred sleep-wake rhythms, and in principle an ESM period can start when sensors detect a change in functioning. Because triggered moments are not a random sample of the day, designs usually keep a baseline time-contingent schedule alongside triggers. Battery drain, extra devices and privacy unease add burden. Unlike passive sensing beside ESM, sensing here schedules the prompts.

Attributes

Loss type
Siloed
Data
Mixed
Effort
Weeks

Steps

  1. Decide whether you need event coverage, opportune timing, or both.
  2. Choose sensors and detection rules (thresholds or classifiers) and validate them in a pilot against ground truth.
  3. Program triggers alongside a baseline time-contingent schedule to preserve representativeness.
  4. Adapt daily schedules to inferred wake times if needed, and suppress prompts in unsafe contexts such as driving.
  5. Track trigger accuracy, compliance, response delay and burden (battery, extra devices).
  6. Analyse with awareness that triggered moments are not a random sample of the day, comparing triggered and scheduled prompts.

Sources

  • The Open Handbook of Experience Sampling Methodology (Myin-Germeys & Kuppens, 2022)
  • Intensive Longitudinal Methods (Bolger & Laurenceau, 2013)
  • Experience Sampling Method (Rhee et al., 2020)

Also known as

  • Context-aware experience sampling
  • Sensor-triggered prompts
  • Opportune-moment prompting
  • Device-contingent sampling

Categories

  • User Research
  • Field & Ethnographic