While it
is probably the case that some individuals were
recorded more than once, given the relatively short
time periods measured at separate buildings, the
number of redundant counts is likely to be only a small
proportion of the total ridership and likely to be distributed nonsystematically across conditions.