Composite Designs in Agricultural Research

Robinson, Philip J.; Nielsen, K. F.

doi:10.4141/cjss60-022

Cited by 5 publications

(5 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These conditions are often fulfilled satisfactorily in agronomic investigations as indicated by the fact that in six different studies conducted by the authors the fitted surfaces passed close to the treatment averages. This has also been the experience of others ( 1, 2, 9, 10,11,12,13,14).…”

mentioning

confidence: 59%

“…In contrast, information from ordinary factorial data estimates the interaction coefficients with greater precision than the pure quadratic coefficients. Although polynomials may be fitted to factorial data ( 14), the more convenient calculation situation produced by the "Box" designs is often very important to the agronomist who does not have ready access to computer services.…”

mentioning

confidence: 99%

See 1 more Smart Citation

An Agronomically Useful Three‐Factor Response Surface Design Based on Dodecahedron Symmetry¹

Hermanson¹,

Gates²,

Chapman³

et al. 1964

Agronomy Journal

View full text Add to dashboard Cite

Synopsis This design allows application of 3 quantitative independent variables at 7 preselected levels each and the fitting of a second degree polynomial to the quantitative dependent variable data. Thirty plots per replicate can be divided into 5 equal sized blocks of 6 plots. An estimate of error with five degrees of freedom is available within each replicate. This design has proven practical for soil fertility field experimentation and an example of this application is given.

show abstract

mentioning

confidence: 59%

mentioning

confidence: 99%

An Agronomically Useful Three‐Factor Response Surface Design Based on Dodecahedron Symmetry¹

Hermanson¹,

Gates²,

Chapman³

et al. 1964

Agronomy Journal

View full text Add to dashboard Cite

show abstract

“…Stated otherwise, this is a test of the planar response surface followed by a test of the improvement in fit of the equation to the data when curvature terms are introduced. Robinson and Nielsen (1960) contested this approach and suggested that all orthogonal (b1, b2, b3, b12, b13, b23) terms be tested individually. The latter approach was employed in this work because of an interest in detecting and inspecting interactions.…”

Section: Resultsmentioning

confidence: 99%

“…Central composite design is particularly suited to experimentation with material of low variability where a sequential approach is possible. Several writers have used these designs for the study of biological problems, among which the papers of Hader et al (1957), Moore et al (1957), Miller and Ashton (1960), and Robinson and Nielsen (1960) are cited.…”

Section: Selection Of Experimental Designmentioning

confidence: 99%

Statistical Measurement of Biotin, Thiamine, and Zinc Concentrations Required for Maximal Growth of Arthrobotrys conoides1

Grant¹,

Coscarelli²,

Pramer³

1962

Applied Microbiology

View full text Add to dashboard Cite

Brunswick, N.J.), WALDIMERO COSCARELLI, AND DAVID PRAMER. Statistical measurement of biotin, thiamine, and zinc concentrations required for maximal growth of Arthrobotrys conoides. Appl. Microbiol. 10:413-417. 1962.-An experiment of central composite design was performed to determine simultaneously the most suitable concentration of each of three nutrients essential for proliferation of the nematode-trapping fungus, Arthrobotrys conoides, and to detect any significant interaction that occurred when they were supplied at different levels. Regression analyses of the results obtained revealed that for maximal growth of A. conoides in a glucose-inorganic salts medium, biotin, thiamine, and zinc were required at concentrations of 5, 100, and 400 ,ug/liter, respectively. No interaction of significance was detected. The analytical procedures are presented in detail, and the benefits that accrue from a statistical approach are discussed. Elucidation of the nutritive requirements of a biological system consists of two principal aspects. These can be categorized as: identification of those factors essential for growth, and determination of concentrations necessary for maximal development. The present paper is concerned with the latter aspect only: specifically, the optimal concentrations of biotin, thiamine, and zinc which were demonstrated as required for growth of the nematodetrapping fungus, Arthrobotrys conoides, in a glucoseinorganic salts medium (Coscarelli and Pramer, 1962). Traditionally, in a study of this type, a series of experiments is conducted and each factor is varied in turn while holding all others constant. Although this is a valid method of investigation, too frequently it is not realized that the results obtained are applicable only under the specific conditions employed in the tests. Use of designed experiments, involving the simultaneous variation of more than

show abstract

“…Among the more important are plot size and experimental design (for example, a composite ora factorial design [10]). The number of factor levels used will also influence estimated process variance [2, p. 170].…”

Section: Research Costsmentioning

confidence: 99%

Economic Considerations in Response Research

Anderson

Dillon

1968

American J Agri Economics

View full text Add to dashboard Cite

The general negleet of economic considerations in response researeh is noted. Following the listing of several factors which influence investment in response research, a framework is developed in whieh knowledge of a process is viewed in terms of the expected value and variance of profits. A Bayesian scheme for incorporating additional information in posterior analysis is outlined. Worth of additional information is judged in terms of utility. The seheme is used to show how the value of obtaining additional information can be aseertained ex ante in preposterior analysis by way of a Monte-Carlo approach. Both posterior and preposterior analyses ate illustrated through a soybeanJertilizer process. In this empirical example, the effects of using different experimental designs of varying size ale examined. F OR A decade or more, agricultural economists (production variety) have been urging agricultural scientists to estimate crop and livestock production functions. Such estimates, we have argued, will enable farmers to make better decisions on input and output levels in their crop and livestock production. At the same time, no overt attention has been given to the economics of conducting such research. No concrete guidance has been given as to the extent to which response processes might best be investigated in terms of costs and returns.Seemingly ah act of professional faith, our evangelizing has implied that production function estimation based on experiments with "many factors at many levels rather than many replications of a few factors" will give information of greater value than its cost. Or else, in chasing the MVP of the ith factor, we have ignored the MVP of response research itself. It is as ir the mere organization of data in a form aesthetically suited to applying production economics principles has been an end in itself. Either of these implications leaves much to be desired. Here we hope to make a start towards rectifying this situation by providing some appreciation of the economics of response research.To make the problem tractable, we confine our attention to the situation of a farmer interested in investing in his own response research. Thus, by taking the problem out of the public arena, we avoid the peripheral complications of, on the cost side, having to consider fiscal policies for the financing of public research, and interrelatedly, on the revenue side, having to assess the aggregate social benefits of public response research. Not that these macro problems are unimportant, but they can hardly be appraised without first having ah appreciation of response research in its micro aspects.* With the usual caveat, we are indebted to John Phillips for critical comments.J. R. ASDERSOI~ AND ffOItN L. DIIa~s ate, respectively, research feUow and profe, sor in the

show abstract

Composite Designs in Agricultural Research

Abstract: An experiment was conducted in the greenhouse using tornato plants for tesdng

Cited by 5 publications

References 3 publications

An Agronomically Useful Three‐Factor Response Surface Design Based on Dodecahedron Symmetry¹

An Agronomically Useful Three‐Factor Response Surface Design Based on Dodecahedron Symmetry¹

Statistical Measurement of Biotin, Thiamine, and Zinc Concentrations Required for Maximal Growth of Arthrobotrys conoides1

Economic Considerations in Response Research

Contact Info

Product

Resources

About

Composite Designs in Agricultural Research

Abstract: An experiment was conducted in the greenhouse using tornato plants for tesdng

Cited by 5 publications

References 3 publications

An Agronomically Useful Three‐Factor Response Surface Design Based on Dodecahedron Symmetry1

An Agronomically Useful Three‐Factor Response Surface Design Based on Dodecahedron Symmetry1

Statistical Measurement of Biotin, Thiamine, and Zinc Concentrations Required for Maximal Growth of Arthrobotrys conoides1

Economic Considerations in Response Research

Contact Info

Product

Resources

About

An Agronomically Useful Three‐Factor Response Surface Design Based on Dodecahedron Symmetry¹

An Agronomically Useful Three‐Factor Response Surface Design Based on Dodecahedron Symmetry¹