Agnes Scott College

Gertrude M. Cox and William G. Cochran

Experimental Design
John Wiley & Sons (1st Edition, 1950; 2nd Edition, 1957)
Wiley Classics Library, reprint of 2nd Edition, 1992

Experimental Design


Since statisticians do not usually perform experiments, their claim to attention when they write on this subject requires some explanation. It is true that on many important aspects of experimentation the statistician has no expert knowledge. Nevertheless, in recent years, research workers have turned increasingly to statisticians for help both in planning their experiments and in drawing conclusions from the results. That this has happened is convincing evidence that statistics has something to contribute....

...From this point of view, the problem of summarizing the results may be restated in the question: what can we say about the true difference between A and B? This is a problem in induction from the part to the whole, or in statistical language, from the sample to the population. A solution to this problem has been developed by means of the theory of statistics. It is this solution that constitutes the principal contribution of statistics to the interpretation of the results.

Table of Contents

  1. Introduction
  2. Methods for Increasing the Accuracy of Experiments
  3. Notes on the Statistical Analysis of the Results
  4. Completely Randomized, Randomized Block, and Latin Square Designs
  5. Factorial Experiments
  6. Confounding
  7. Factorial Experiments in Fractional Replication
  8. Factorial Experiments with Main Effects Confounded: Split-Plot Designs
  9. Factorial Experiments Confounded in Quasi-Latin Squares
  10. Some Methods for the Study of Response Surfaces
  11. Incomplete Block Designs
  12. Lattice Designs
  13. Balanced and Partially Balanced Incomplete Block Designs
  14. Lattice Squares
  15. Incomplete Latin Squares
  16. Analysis of the Results of a Series of Experiments
  17. Random Permutations of 9 and 16 Numbers