Precision of Field Experiments With Vegetable Crops as Influenced by Plot Size and Shape: Iii. Potatoes

Abstract

uniformity trial data together with certain cost assumptions fof potatoes indicate thaf long narrow p-iots are-the mosr efficienr. The basic unit was 4 tt io"" ""a 3 ft wid"e. A minimum of rwo basic units would be necessary for most p"i?tr .t"ai.t with a maximum of slx basic units-depending.upon the,requiremenrs for each experiment, Thus the optimum plot size co.uld range from 3 tt ;ft;-;a a to ts i'c long' Where smallcr size becomes optimum more replicaaio"r "t" tuq"i*a. fotig narrow replicates likewise appeir to be more elficient than either wide or square ones'The loeical analysis and a generalized flow diagram for digital computer calculatio"n of these data are presented.

Estimates of Optimum Plot Size and Shape and Replicate Number for Forage Yield of Alfalfa‐Bromegrass Mixtures1

“…The average estimates of C 1 and C 2 made by 10 agronomists were 70% and 30%, respectively. These cost estimates were used in the cost function [ 2] to compute the relative cost in obtaining treatment means of equal accuracy for the various plot and block combinations. The estimated relative cost of plots larger than the basic unit was higher than the cost of the basic unit plots for all block shapes with one exception.…”

“…Smith's method and the coefficient of vanat10n method gave similar estimates of optimum plot size for sweet corn ( 2). Plot efficiency was influenced by both plot and block shape.…”

“…The coefficient of soil heterogeneity was computed as the regression of the logarithm of the variance of yield per basic unit (Vn) on logarithm of size of plot in basic units (n) using the among plot variance (no blocking) of the 12 sizes and shapes of plots (Table 1) as described by Smith ( 4) and Steel and Torrie ( 5). When the cost is given by [2] and the variance in form of [3] the total cost for a given variance is a minimum if plot size is determined by Nopt= bCd(1-b)Cz [4] where C,, C" and b are as defined before and N •• , is the optimum plot size. The size of the difference, d, that can be detected in percent of the mean at the 5% level of significance and with 80o/o assurance was estimated for the different sizes and shapes of plots for the replication number given in Table 2 for the 1-plot X 12-plot block shape by utilizing expressions [1) and [3)…”

“…For plots of 1, 2, 3, and 4 units long and 1 unit wide the relative precision of incomplete blocks as compared to complete blocks was determined by procedures given by Cochran and Cox ( 1). The required number of replicates and the relative cost of the 4 plot and the 2 block sizes were determined by expressions [ 1] and [ 2].…”

People, Languages, and Computers: A Short Memoir

“…I contributed the Appendix, which included a flow diagram with the comment that it was intended to be "sufficiently general so that it may be used in preparing the program in any other programming system for other computers," a statement I find now to be both embarrassing and amusing. 5 I believe now that flow diagrams are a reflection of the inadequacy of many computer languages to represent algorithms. Indeed, since using the array languages I discuss later, I cannot recall having prepared a flow diagram.…”