Missing and Partial Datasets

Top reasons it may appear there is partial data, when all data is there:

  • A mobile user hasn’t responded to all sessions, or completed all sessions or questions.
  • When using branch logic, each mobile user won’t see every question in a given session. The reason is whether they see certain branching questions depends on their responses.  This can make it seem data is missing for certain mobile users, when it really isn’t.
  • In wide format data, each row represents a session and each column represents a different variable within a session. If you create different sessions, only the questions for a specific session will have data. For example, if you created a morning session with one set of questions and an evening session with another set, a row should only show data for questions from one session (e.g., morning session) and the columns (questions) for the other (e.g., Evening) session should be blank for that row.
  • GPS data may be blank if the mobile user has turned off their location service or cannot access a signal.

 

Top reasons for partial data:

  • The mobile user doesn’t have an internet connection while the mobile app is open. This can create a lag in data being uploaded.
  • The mobile user hasn’t interacted with the mobile app for an extended period of time, or at the end of a study.

 

______________

 

Our mobile app collects 4 types of data:

  1. NIS data is related to when notifications are received, whether a mobile user responds and when.
  2. Session response data is related to whether a session is completed and when.
  3. Prompt response data includes a mobile user’s response to questions, whether they viewed a prompt and answered it, and how long they took.
  4. Reminder data is related to how many reminders were sent before a mobile user responded to a notification.

 

Data is sent from the mobile app to the server at 4 different times: (All of these require an internet connect)

  1. When the mobile app is brought from the background to the foreground. (Data type: NIS data, Reminder data)
  2. When a session starts. (Data type: Session response data, Prompt response data)
  3. When a session ends. (Data type: Session response data, Prompt response data)
  4. When the “My Lifepak” page is up. (Data type: NIS data, Reminder data)
  5. When the “Manual Data Upload” feature is used. (Data type: all)

 

A few things to note:

  • If the mobile app is in closed or in the background, it cannot send data to our server. This is a limitation with the Android and iOS operating systems. We have designed the app to send data whenever possible, while also trying to maximize the user experience and avoid delays that frustrate users (and reduce engagement).
  • If a mobile user doesn’t have an internet connection during a study, or toward the end of a study, their data might not have had a chance to upload to the server. Their data will upload once they have an internet connection and interact with the mobile app.
  • If a mobile user stops interacting with the mobile app, some data may not have a chance to upload to the server. This can happen for data toward the end of a study, if mobile users stop responding.  If they interact with the mobile app, and have an internet connection, then this data will be uploaded.
  • Even if a participant deletes a LifePak, the mobile app will save the data and should upload it the next time the mobile user opens the app.
  • Data can always be manually uploaded through the “Manual Data Upload” feature of the mobile app (under “Settings”). (Note: If you are having participants come back at the end of the study, you might have them bring their mobile devices.  This way you can ensure all data has been uploaded using the “Manual Data Upload” feature before they delete the app.)
Have more questions? Submit a request

Comments