Statistical Methods
When a series of measurements is made in any research, they will not all agree, because the work must inevitably be affected by experimental errors, and it may also be affected by other influences which cannot be eliminated, or which can only be partly eliminated, by the experimental technique available. Even if one or other of these influences can be eliminated, the value of the research may be reduced by the restriction, because then no information is provided on the action of the eliminated
influence, which may be of importance.
J.J. Leeming
Author(s): Leeming, J J