To calculate the mean, what you should do is add up all the values, and divide by how many there are.
What you don't do is, on Aug 8, you don't add up the eight values you have and then divide by 31.
Just in case the list of chart bugs wasn't long enough.
Also, that 6h 44m average across the calendar YTD: that isn't right, either. The true picture is that every month so far is above 7h 15m. For the chart shown, either you're averaging the monthly averages (oops - another class missed in stats 101), or repeating the first mistake and counting to 31 August. Or doing a third crazy thing, which on current performance is quite likely.
Really? Is this Fitbit's level now? And yet, you've got 510(k) clearance from the US FDA and CE marking in the EU for your ECG app for AFib. The maths required for that is advanced and it matters if you get it wrong.
At the same time you're losing people's fitness data: exercise and weight records frequently vanish, as noted in this forum and elsewhere, which is an actionable violation of EU (and UK) GDPR.
I don't think the present situation looks good for Fitbit, honestly.
Best AnswerOMG it's worse than I thought. According to the monthly chart, April was 7h 26m:
But according to the calendar YTD chart, it was 6h 56m:
So which is it?
I haven't got the time to download the raw data and calculate it for myself.
How anybody has any faith left in any of the numbers your app displays I have no idea.
Best Answer