The Design of Experiments

Chapters

1. Introduction
2. The principles of experimentation, illustrated by a psycho-physical experiment
3. A historical experiment on growth rate
4. An agricultural experiment in randomized blocks
5. The Latin square
6. The factorial design in experimentation
7. Confounding
8. Special cases of partial confounding
9. The increase of precision by concomitant measurements. Statistical Control
10. The generalization of null hypotheses. Fiducial probability
11. The measurement of amount of information in general

    Quotations regarding the [null hypothesis]

• "...the null hypothesis is never proved or established, but is possibly disproved, in the course of experimentation. Every experiment may be said to exist only in order to give the facts a chance of disproving the null hypothesis."
• "...the null hypothesis must be exact, that is free from vagueness and ambiguity, because it must supply the basis of the 'problem of distribution,' of which the test of significance is the solution."
• "We may, however, choose any null hypothesis we please, provided it is exact."

• "For this purpose the new test proposed would be entirely inappropriate, and no experimenter would be tempted to employ it. Mathematically, however, it is as valid as any other, in that with proper randomisation it is demonstrable that it would give a significant result with known probability, if the null hypothesis were true."

• "The notion that different tests of significance are appropriate to test different features of the same null hypothesis presents no difficulty to workers engaged in practical experimentation, but has been the occasion of much theoretical discussion among statisticians."

• "This question, when the answer to it is not already known, can be fruitfully discussed only when the experimenter has in view, not a single null hypothesis, but a class of such hypotheses, in the significance of deviations from each of which he is equally interested."