The data below are for the time taken (in minutes) for people to evacuate a multi-storey building in response to a fire alarm that was tested on two consecutive days. In both cases, there was no fire. A human movement scientist hypothesised that experiencing a false alarm on the first day would affect the time a person took to evacuate the building on the second.
(a) What sort of statistical analysis is appropriate for this hypothesis?
(b) Is the hypothesis one- or two-tailed?
(c) Is the result of the analysis significant?
(d) What does this suggest about the behaviour of people in response to fire alarm testing?