Abstract
ONE of the striking characteristics of modern biological statistics is its insistence on experimental design. The most powerful methods of analysis can be used, and the greatest amount of information obtained, only if the experiment is laid out with this end in view. Randomization is necessary for trustworthiness, replication for efficiency and factorial design for comprehensiveness. Without these precautions, the experiment will at best give a poorer return than it need, for the labour and materials expended on it.
The Design of Experiments
By Prof. Ronald A. Fisher. Fourth edition. Pp. xi + 240. (Edinburgh and London: Oliver and Boyd, 1947.) 12s. 6d. net.
Article PDF
Rights and permissions
About this article
Cite this article
M., K. The Design of Experiments. Nature 160, 775 (1947). https://doi.org/10.1038/160775a0
Issue Date:
DOI: https://doi.org/10.1038/160775a0