> You use them over long periods of time to either get aggregate statistics to compare against baseline or track personal changes over time. Doesn't that assum

> You use them over long periods of time to either get aggregate statistics to compare against baseline or track personal changes over time. Doesn't that assume that the errors will be consistent and precise enough that any trends don't get totally wiped out?