An unmonitored participant turns on the iPhone app, walks in a "straight line", turns around and "rests", and then "walks back in a straight line" with the phone in a pocket (front or back) on left or right side, or maybe in a bag? Is this a correct summary of the "protocol"? I assume they are asked to walk on a "flat surface".
So if a user doesn't follow the protocol well, is it used as part of the scoring/leaderboard scenario? Are you defining "truth" based on bad events?
Created by Monte Shaffer MonteShaffer Hi Monte:
We had to do a little digital archeology - there have been 29 different versions of the app. And although most of them used overlapping instructions we have found at least three different ones:
1. The one in the video - walk for 20 steps (but timer is also running for 30s) turn around and wait for 30s then walk back
2. walk for 20 steps (but timer is also running for 30s) turn around walk back for 20 steps (also with 30 s timer) followed by standing for 30s
3. Walk for 30s "back and forth" then turn around 360 degrees then stand still for 30s
It is possible that there are small changes in between these ones as well. Did the protocol instructions change over time?
Based on a few observations of timestamps (stitching a record together: outbound,rest,return) ... I seem to be getting outbound, then return then rest frequently.
${imageLink?synapseId=syn10734072&align=None&scale=100&responsive=true}
It is unclear if your "image-sharing" above is private. That is correct Solly, in version 1.3 and above the instructions say to walk for 30s then turn around 360 degrees then stand still for 30 seconds. In the released data the instructions are as @MonteShaffer summarizes.
You can see a screen cast of the instructions here: https://www.synapse.org/#!Synapse:syn4993293/wiki/376006. This does not play the full standing test, nor the return. I believe in version 1.3 and beyond, there is no inbound walking test. @larssono, correct me if I'm wrong. I don't see any instructions about walking on a flat surface.
Since these data are unmonitored/self-reported, there may be varying degrees of non-compliance and/or artifacts in them. It's definitely something to keep in mind as you analyze them.