No interesting experiment gives a precise and unambiguous result; there is always a margin of error. For example, if you sample 100 people out of the workforce of a large company, by chance you might choose those who are unusually tall or short, or those who are unusually happy or unhappy with the systems they use. Furthermore, there may be limits to the accuracy of the measurement devices you use, although this is more of a problem in physical rather than human sciences. Where the error is due to random effects, and these are understood, you may be able to estimate the margin of error, for example the standard error (s.e.) of the mean.
Used on page 65