Full Sensing vs Sensing Lite – Which is best for your research?

If you want to use mobile sensing in m-Path, you have the choice between getting a full sensing setup (with m-Path Sense), or using the limited sensing capabilities in the regular m-Path app. The main difference is that, in sensing lite, you can only get sensor measurements during a questionnaire. In sensing full, it possible to constantly get sensor measurements. On this page, we will provide a quick overview of both possibilities.

Full Sensing

The full sensing set-up requires your participants to download a separate m-Path app, named m-Path Sense. This app is practically equivalent to the regular m-Path application, with the only difference being that it will constantly collect sensing data in the background. This data will then be periodically uploaded to your selected cloud storage location. Full sensing essentially means two things: firstly, that the sensing is continuous, so the sensors are recording data even when the app is in the background. Secondly, the app tries to use as many sensors as possible, limited only by the hardware or the permissions set by the user. These include:

• App Usage and Screen Activity data
• Accelerometer, Gyroscope and Light sensor data
• Battery level data
• Bluetooth and Wi-Fi data (cryptographically anonimized)
• GPS location data
• Pedometer and Physical activity data
• Call and text data (cryptographically anonimized)

The upside of this approach is that you get a very complete image of the activities of your participant. Since smartphones have become a central part of the daily routine, knowing everything a phone goes through throughout the day gives a pretty accurate picture of what the phone’s owner experiences. Combined with classical ESM surveys, you have a lot more data to reason with.

However, there might be downsides. For starters, participants might not like how invasive this sensing approach feels. Since the sensing is continuous, on first startup the participant has to allow all the above sensors to be tracked, which could throw them off. As such, always be sure to discuss full sensing research thoroughly with your participants. Also, full sensing tends to drain device batteries more quickly, which makes daily phone charging necessary, or even charging throughout the day for older devices. Both the iOS and Android system will try to kill apps that are not actively used and are draining battery. Therefore, the m-Path Sense app needs to be restarted occasionally. Last but not least, running and setting up a full mobile sensing study can be quite intensive. If these disadvantages seem daunting, see if limited sensing (aka Sensing Lite) would work for you.

Sensing Lite

In the regular m-Path app, you can also use some sensing capabilities, given you purchase the Mobile Sensing Lite feature pack. These work differently however. While the full sensing set-up constantly captures sensing data, sensing in the regular m-Path app works similarly to asking a question in a survey: by adding a sensing module in your survey. This works as you would expect: when the user fills in a survey and reaches a sensing question, the app will request the relevant sensor data at a single point, and store it as an “answer”.

The upside of this approach is that you have great control over when and where to ask for sensing data. You can choose to only capture it during specific surveys, and even then you can use branching logic to only capture it in specific cases. This also mitigates the battery impact of the sensing, and in this case, participants are only asked to give permission when the app actually asks for the data.

The downsides are that you have a more limited view of the sensor data, since they’re only momentary data points, and if the client neglects filling in the survey, no sensor data is collected. The types of data you can collect like this are also limited to:

• GPS location (at this moment)
• App usage data (of the last day)
• Step count data (since last boot)
• Noise level (at this moment)
• Nearby bluetooth devices (at this moment)
• Audio recording (as an answer to a question)

Conclusion

So, which sensing solution is right for you? In the table below we quickly sum up the most important differences between the two methods, so depending on your use case, you can decide which one works best for you.

	Full Sensing	Sensing Lite
Sensing options (detailed above)	Many	Limited
Sensing records	Continuous	At survey time
Battery impact	Medium	Practically zero
Permission request	At first boot	When needed
Price	Higher	Lower
Difficulty to run and set up	Medium	Low