mEMA10 Sensor Data Guide
The mEMA10 app passively collects sensor data from the participant's smartphone at regular intervals (approximately every 10 seconds when the app is active). This data provides objective, contextual information about the participant's environment and phone usage patterns without requiring any manual input.
All sensor readings are timestamped and uploaded automatically alongside survey responses.
Sensors collected
The following table describes each sensor, the range of values it can produce, when readings are collected, and what those readings represent.
Sensor | Possible values | When collected | Description |
Acceleration | 0.000 to ~1.5+ | Every ~10 seconds, when enabled, app is in foreground, and sufficient movement is detected | Measures the magnitude of device movement. 0 = stationary; higher values indicate more vigorous movement, such as walking or handling the phone. Collection is limited to foreground use only — iOS reserves background motion data for native system apps. |
Max Light | 0.000 to 1.000 | Every ~10 seconds | The highest ambient light value recorded during the sampling window. Captures brief moments of brightness even if overall conditions are dim. |
Min Light | 0.000 to 1.000 | Every ~10 seconds | The lowest ambient light value recorded during the sampling window. Captures brief moments of darkness even if overall conditions are bright. |
Average Light | 0.000 to 1.000 | Every ~10 seconds | The mean ambient light level across all samples in the window. Provides a smoothed picture of the overall lighting environment. |
Screen On | 0 or 100 (%) | Every ~10 seconds | Percentage of the sampling window during which the phone screen was on. 100 = screen was on the entire window; 0 = screen was off the entire window. |
Screen Locked | 0 or 100 (%) | Every ~10 seconds | Percentage of the sampling window during which the phone was locked. 100 = phone was locked the entire window; 0 = phone was unlocked. |
App Visible | 0 to 100 (%) | Every ~10 seconds | Percentage of the sampling window during which the mEMA app was in the foreground. 100 = app was fully visible; 0 = app was in the background. |
GPS (Latitude / Longitude) | Lat: -90 to 90 Lon: -180 to 180 | When location is enabled in the server configuration, and the participant has accepted the location permission | The geographic coordinates of the participant's device. A reading of 0 indicates no location was available, not that the participant is at that location. |
Sensors not collected
Several sensors that may have been available in the past but are not collected in mEMA10. The table below explains why each has been disabled.
Sensor | Reason not collected |
Proximity | Enabling the proximity sensor causes the phone screen to turn off when the sensor is covered — the same behaviour seen during a phone call when the screen goes dark near the ear. This would interfere with survey completion and could distort responses, particularly during cognitive tasks. |
Orientation | The mEMA app is locked to portrait mode, so orientation data does not change meaningfully. Additionally, orientation can only be captured while the app is active in the foreground, making background collection impossible. |
Amplitude (microphone) | Collection of audio amplitude data is restricted under Apple's privacy framework and cannot be accessed by third-party apps. |
Step count | Step count data is collected more accurately through the participant's Garmin watch, which is dedicated to activity tracking. Phone-based step counting is therefore redundant and has been disabled. |
Activity type/confidence | These fields are derived from step count data. Because step count is disabled and collected via Garmin instead, activity type and confidence values are not available. |
Humidity, temperature, barometric pressure | Modern smartphones no longer include environmental sensors for humidity, temperature, or barometric pressure. These sensors are not available on participant devices and cannot be collected. |
Configuration requirements
Two sensors require additional setup before data will be recorded:
Acceleration — must be enabled on the server (mEMA10) configuration page. Data is only recorded when sufficient device movement is detected, and the app is in the foreground. Stationary periods will report 0.
GPS (Latitude / Longitude) — must be enabled on the server configuration page, and the participant must accept the location permission on their device. If either condition is not met, coordinates will be reported as "0, 0".
Data interpretation examples
Individual sensor values are most meaningful when read in combination. The following examples show how to interpret common patterns.
Phone usage patterns
Combining Screen On, Screen Locked, and App Visible lets you reconstruct how participants are interacting with their phone at a given moment.
Time | Screen on | Screen locked | App visible | Likely interpretation |
16:54:07 | 100 | 0 | 90.9 | Phone is on and unlocked with the mEMA app in the foreground — participant is likely completing a survey. |
16:56:27 | 0 | 100 | 0 | Phone is locked, and screen is off — participant has put the phone in their pocket or set it down. |
17:18:36 | 100 | 0 | 0 | Screen is on and unlocked, but the mEMA app is not visible — participant is using another app. |
Activity levels
Acceleration values can help distinguish sedentary periods from active ones, and can supplement self-reported activity data from surveys.
Time | Acceleration | Screen on | Likely interpretation |
16:54:07 | 0.000 | 100 | Phone is stationary with the screen on — participant is likely sitting and using their phone, such as completing a survey. |
18:26:15 | 1.303 | 100 | Significant device movement with the screen on — participant may be walking while looking at their phone. |
20:42:03 | 0.000 | 0 | No movement and screen off — the phone is resting on a surface; participant may have set it down for the night. |
Environmental context
Light sensor values combined with time of day can indicate where participants are and what they might be doing — for example, identifying potential sleep periods or indoor/outdoor transitions.
Time | Max light | Min light | Avg light | Likely interpretation |
20:30:18 | 0.200 | 0.000 | 0.003 | Very low light overall with brief flickers — participant is likely in a dark room with a TV or nearby screen providing occasional light. |
21:36:38 | 0.700 | 0.000 | 0.354 | Moderate average light with a wide range — could be a dimly lit room with variable lighting, or movement between rooms. |
21:18:36 | 0.050 | 0.050 | 0.050 | Consistently very low light with no variation — participant is in a uniformly dark environment, such as a bedroom at night. |
Notes
The sample_count field in the raw data indicates how many individual samples were aggregated into each row. A value of 1 typically corresponds to a state transition event (e.g., screen locking) rather than a regular 10-second sampling window.
GPS coordinates of 0.000000, 0.000000 indicate that no location fix was available, not that the participant is physically at that location.
The location_time field shows when the GPS reading was last updated. It may differ from the row timestamp if the GPS has not refreshed recently.