A small, unpublished reliability study conducted by our group found that the simulated firefighting tasks gad a relatively high level of error (mean typical error=16.9%), as task validity was prioritised over reliability in order to maximise the transferability of results to the fireground