How much sampling bias is there in this, though? The people that own devices that can count steps will tend to be those who are active enough to consider monit

How much sampling bias is there in this, though? The people that own devices that can count steps will tend to be those who are active enough to consider monitoring them something that might be worth doing. And then there are people like my wife that typically get over 5,000 "steps" because she's always doing things with her hands. Her insurance was giving her $60/year for wearing my old FitBit (I had upgraded to the waterproof version.) Now it's down to $20/year and soon she's going to quit wearing it. What they will really be measuring is that $20/year is not going to prompt her to replace it when the battery finishes dying.