Got a bigger sample?

Jim Foulds, chair of the Higher Education Safety Forum, takes The Times Higher to task for reporting a study of university stress based on a sample of "only" 1,000 respondents (Letters, February 9). If he has access to data from a larger sample, yielding different findings, why doesn't he say so? Otherwise, what's wrong with a sample size of 1,000?

As for Foulds's claim that the survey "related to term time only, not to averages over the whole year", this may be because terms are the relevant units of working time. Heaven help any of his family members if they report they have a temperature of 40C and get told this means nothing averaged over a lifetime.

Roger Lindsay.
Ulverston, Cumbria