research best practice

[jpoyser]

I am undertaking a research thesis using the Charge 4 with 30 participants. I have had to go through an ethics process ensuring the privacy of the participants is very protected.

I am able to connect and extract data for a participant using the web api.

Is there a best practice page or article for research setup? 

ie should I set up the users fitbit accounts before I provide the wearables to the research participants?

If I don't want participants to use their own accounts for privacy reasons is there a way to setup multiple accounts? I'm currently finding limitations creating gmail accounts.

Is the research participants detailed data collected immediately and then is it just a case of creating my app and getting permission from the participants to retrieve the data from the web through the API's?

Does the data collected during the research process remain indefinitely or is it archived at some point? If I want the data to remain until I've finished analysis do I need to let anyone know?

Any advice or learnings/risks from others would be greatly appreciated.

Regards

Justin

[Gordon-C]

Hi @jpoyser 

The Fitbit Developer Support team doesn't have any best practices specific to research use cases.   Most researchers have their own process for onboarding participants.   We just require that you do not violate our platform terms of service, https://dev.fitbit.com/legal/platform-terms-of-service/.  I do have some other links from our Health Solutions team that might give you some guidance.

Research Pledge - https://healthsolutions.fitbit.com/research-pledge/

Researchers FAQ - https://healthsolutions.fitbit.com/researchers/faqs/

Some of the technical questions I can answer

1. The participants data is collected when the user syncs their device with the Fitbit mobile application.   The sync will automatically occur about every 15 minutes if the mobile application is running on the phone, the phone has a data connection, and the device is within bluetooth range.  Once the sync is finished, the data is available immediately.  

2. Once the user consents to share their data with your application, you have access to their Fitbit data.

3. The data is available as long as the participant maintains consent with your application and doesn't delete their data.   We do not archive the data.

There are some researchers who follow the forums, so maybe one of them can provide some suggestions for you.

Gordon

[cwebster_nymbly · Accepted solution]

Hi @jpoyser - 

I've used various wearables for clinical research and here's the workflow we follow:

1. Participant consents to the research study
2. Participant is provided their device (either shipping to them or they can pick it up, depending on how your study is designed)
3. Participant sets up their device independently (e.g., they create their own account within Fitbit, get the app, etc)
4. Participant uses oauth to allow data sharing
5. You use the API to collect their data based on their study participation & your study design
6. Once the subject has completed or withdrawn from the study, you make sure you've collected the data and then you revoke access

There may be an additional step where you re-collect the devices, but we've typically allowed the participant to keep them as a research incentive so we simply revoke access after we've grabbed the study data. 

Hope that helps!

- Courtney

[jpoyser]

Thanks for the responses @Gordon-C  and @cwebster_nymbly. I ended up creating the accounts myself (I used gmail with a + after one email address and it then lets me create as many email addresses as I want that all link to the one account) I also created the fitbit accounts too and signed the users up myself. That way I can keep the data and information anonymous throughout the study and as the cohort are all over 60 I can ensure the device is working, their syncing their data etc. Lots of lessons learnt going through the process that is for sure.

I'm also using the first week as a flesh the process out before the real study begins. I think there would be merit in creating a forum or a pinned post on best practice for starting out... lots of great information in these forums too.

Cheers and thanks again for your advice.

Justin