University of California

Biases encountered in large-scale yield tests


O. C. Riddle
G. A. Baker

Authors Affiliations

O. C. Riddle was Instructor in Agronomy and Junior Agronomist in the Experiment Station. Formerly Agent, United States Department of Agriculture; G. A. Baker was Assistant Professor of Mathematics and Assistant Statistician in the Experiment Station.

Publication Information

Hilgardia 16(1):1-14. DOI:10.3733/hilg.v16n01p001. February 1944.

PDF of full article, Cite this article


Abstract does not appear. First page follows.


Certain difficulties in interpreting the results of the ordinary analysis of variance when applied to yield tests of genetically similar wheat strains, derived through backcrossing, prompted a critical examination of the data. These data indicated a bias inherent in any large-scale experiment that imposes an inflexible design upon a soil whose fertility may fluctuate markedly within short distances. Because of this bias, the usual analysis of variance tends to overestimate significance grossly when the number of varieties is large and the productivity levels of the soil change rapidly and erratically (as at Davis). One may briefly describe the bias by saying that the inflexible design of the experiment tends to subtract too little from the naturally high-yielding plots and too much from the naturally low-yielding plots in attempting to correct for differences in soil productivity. This has a spreading effect on the part of the variation that is labeled “varietal differences” and thus causes a serious overestimation of significance.

In a conventional analysis of variance, the natural variation of an experiment is arbitrarily partitioned into categories according to a preconceived mathematical model. These categories are labeled “variation due to varieties,” “variation due to soil productivity,” and so on. If the experiment is in exact accord with the model, the labels are accurate. If the experiment is not as called for by the model, the labels are misleading: for instance, the category labeled “variation due to varieties” may contain some of the variation due to soil-productivity.

A mathematical model may be pleasing and beautiful to its creator or users. If, then, nature does not conform to the model, workers having limited firsthand experience with the vagaries of biological material may even feel that nature has erred and should be corrected.

Literature Cited

[1] Baker G. A. Fundamental distribution of errors for agricultural field trials. Natl. Math. Mag. 1941. 16:7-19. DOI: 10.2307/3028105 [CrossRef]

[2] Baten W. D., Northam J. I., Yeager A. F. Grouping of strains or varieties by use of a Latin square. Amer. Soc. Agron. Jour. 1941. 33:616-22. DOI: 10.2134/agronj1941.00021962003300070004x [CrossRef]

[3] Briggs Fred N. The backcross method in plant breeding. Amer. Soc. Agron. Jour. 1935. 27:971-73. DOI: 10.2134/agronj1935.00021962002700120003x [CrossRef]

[4] Briggs Fred N. The use of the backcross in crop improvement. Amer. Nat. 1938. 72:285-92. DOI: 10.1086/280783 [CrossRef]

[5] Cox G. M., Eckhardt R. C., Cochran W. G. The analysis of lattice and triple lattice experiments in corn varietal tests. Iowa Agr. Exp. Sta. Res. Bul. 1939. 281:1-66.

[6] David F. N. Tables of the correlation coefficient. 1938. London, Eng.: Biometrika Office, University College. 38p. Text Tables 55 p

[7] Fisher R. A. Discussion of: Yates, F. Complex experiments. Roy. Statis. Soc. Jour. Sup. 1935. 2:229-31.

[8] Jones Donald Forsa. Genetics in plant and animal improvement. 1925. New York, N. Y.: John Wiley and Sons, Inc. 568p.

[9] McCarthy M. D. On the application of the Z-test to randomized blocks. Ann. Math. Statis. 1939. 10:337-59. DOI: 10.1214/aoms/1177732146 [CrossRef]

[10] Neyman J. Statistical problems in agricultural experimentation. Roy. Statis. Soc. Jour. Sup. 1935. 2:107-80. DOI: 10.2307/2983637 [CrossRef]

[11] Pope O. A. Efficiency of single and double restrictions in randomized field trials with cotton when treated by the analysis of variance. Arkansas Agr. Exp. Sta. Bul. 1936. 326:1-28.

[12] Salmon S. C. Generalized standard errors for evaluating bunt experiments with wheat. Amer. Soc. Agron. Jour. 1938. 30:647-63. DOI: 10.2134/agronj1938.00021962003000080003x [CrossRef]

[13] Snedecor George W. Calculation and interpretation of variance and covariance. 1934. Ames, Iowa: Collegiate Press, Inc. 96p.

[14] Snedecor George W. Statistical methods. 1938. rev. ed. Ames, Iowa: Collegiate Press, Inc. 388p.

[15] Stringfield G. H., Lewis R. D., Pfaff H. L. The 1940 Ohio coöperative corn performance tests. Ohio Agr. Exp. Sta. Spec. Cir. 1941. 61:1-30.

[16] Student. Yield trials. Hunter, H. Baillière’s Encyclopedia of Scientific Agriculture. 1931. 2:1342-61. Baillière, Tindall, and Cox, London, Eng

[17] Suneson C. A., Riddle O. C., Briggs F. N. Yields of varieties of wheat derived by backcrossing. Amer. Soc. Agron. Jour. 1941. 33:835-40. DOI: 10.2134/agronj1941.00021962003300090009x [CrossRef]

[18] Welch B. L. On the Z-test in randomized blocks and Latin squares. Biometrika. 1937. 29:21-52.

[19] Yates F. Complex experiments. Roy. Statis. Soc. Jour. Sup. 1935. 2:181-223. 243-47 DOI: 10.2307/2983638 [CrossRef]

[20] Zuber M. S. Relative efficiency of incomplete block designs using corn uniformity trial data. Amer. Soc. Agron. Jour. 1942. 34:30-47. DOI: 10.2134/agronj1942.00021962003400010004x [CrossRef]

[21] Zuber M. S., Robinson J. L. The 1940 Iowa corn yield test. Iowa Agr. Exp. Sta. Bul. n.s. 1941. P19:519-93.

Riddle O, Baker G. 1944. Biases encountered in large-scale yield tests. Hilgardia 16(1):1-14. DOI:10.3733/hilg.v16n01p001
Webmaster Email: wsuckow@ucanr.edu