Physical Sciences and Mathematics Commons^™

Ordered Alternatives: A Means Of Improving Power, John E. Boyer Jr.

Conference on Applied Statistics in Agriculture

In analysis of variance settings it is often known that if there are any differences among the means, those differences will fall in a particular order. The usual F test used to look for the existence of differences is not sensitive to the particular order. This paper presents two procedures from the nonparametric literature which have sensitivity to the suggested ordering. The analogy is drawn between these procedures and the two-sample t test. The paper concludes with a simulation study which investigates the power properties of the proposed tests and makes comparisons with the F test.

The Aligned Rank Transform Procedure, James J. Higgins, R. Clifford Blair, Suleiman Tashtoush

Conference on Applied Statistics in Agriculture

Recent work has shown that the rank transform methodology is flawed when applied to multifactor designs with interactions. A simple fix-up is proposed and shown to apply to designs typical of those found in agricultural research including split-plots. Simulation results suggest that the fix-up provides a valid procedure for analyzing multifactor designs when error distributions are symmetric or moderately skewed. The procedure appears to have power advantages over normal theory ANOVA when error distributions are heavy tailed.

Nearest Neighbor Adjusted Best Linear Unbiased Prediction In Field Experiments, Walter W. Stroup

Conference on Applied Statistics in Agriculture

In field experiments with large numbers of treatments, inference can be affected by 1) local variation, and 2) method of analysis .

The standard approach to local, or spatial, variation in the design of experiments is blocking. While the randomized complete block design is obviously unsuitable for experiments with large numbers of treatments, incomplete block designs - even apparently well-chosen ones - may be only partial solutions. Various nearest neighbor adjustment procedures are an alternative approach to spatial variation .

Treatment effects are usually estimated using standard linear model methods. That is, linear unbiased estimates are obtained using ordinary least …

Nonlinear Regression For Split Plot Experiments, Marcia L. Gumpertz, John O. Rawlings

Conference on Applied Statistics in Agriculture

Split plot experimental designs are common in studies of the effects of air pollutants on crop yields. Nonlinear functions such the Weibull function have been used extensively to model the effect of ozone exposure on yield of several crop species. The usual nonlinear regression model, which assumes independent errors, is not appropriate for data from nested or split plot designs in which there is more than one source of random variation. The nonlinear model with variance components combines a nonlinear model for the mean with additive random effects to describe the covariance structure. We propose an estimated generalized least squares …

A Simulation Study Of Field Trial Analysis, Perry Y. Jui

Conference on Applied Statistics in Agriculture

In variety trials, lattice designs are perhaps the most popular ones used by agriculture researchers. An eight by eight lattice design in which there were 56 test cultivars and a check cultivar in each of the eight blocks, was replicated four times. A simulation was performed in which the lattice design was superimposed on two soil fertility maps, one relatively uniform (map 1) and one more heterogeneous (map 2). Ratios of soil variation to total variation (soil + error ) ranging from .1 to 1.0 were studied. The results suggest that in the present setup blocking is more effective when …

Path Analysis In Agricultural Research, K. Bondari

Conference on Applied Statistics in Agriculture

Path analysis introduced by Wright in 1921 as "correlation and causation" has been extensively used in agriculture, sociology, and epidemiology, among many other fields. This study will review path diagrams, algorithms, and the relationship to standardized and mUltivariate regression analyses. Basic assumptions underlying path analysis (e.g., cause and effect relationship, linearity of regression, complete additivity) will also be discussed. Several research examples will be presented to better acquaint statisticians invol ved in agricultural research wi th the methodology and application of path analysis suitable for agricultural data. The method of path coefficient is simple, easy to use, and if "tracing …

A Transformation Approach To Estimating Usual Intake Distributions, Sarah M. Nusser, Alicia L. Carriquiry, Helen H. Jensen, Wayne A. Fuller

Conference on Applied Statistics in Agriculture

Design of effective food and nutrition policies, efficient allocation of resources, and more precise targeting of food programs require good estimates of the percentage of the population with deficient, or excess, nutrient or other food component intake. An individual's mean daily intake of the dietary component is a good estimate of the individual's dietary status. However, to evaluate dietary adequacy of a population it is necessary to obtain an estimate of the distribution of usual intakes. Often, the distribution of usual intakes is estimated from the distribution of mean daily intakes. Two problems arise. First, distributions of usual intakes for …

Linear-Plateau Regression Analysis And Its Application To Selenite Adsorption Rate, Bahman Shafii, Kevin C. Harper, Steven L. Mcgeehan

Conference on Applied Statistics in Agriculture

Simple computational methods are presented which facilitate fitting regression models to response data exhibiting a plateau effect. The iterative statistical program (called PLATFOR) is written in FORTRAN (a SAS version is also available), and produces all relevant regression statistics, plots, and information on goodness of fit. The presented procedures are empirically valuable, since linear-plateau models have many useful applications in agriculture, especially in soil fertility and soil chemistry experiments. The technique was employed in an experiment designed to determine the effect of soil volcanic ash content on selenite adsorption. Ion chromatographic methods were used to investigate selenite adsorption in three …

Applying Principal Component Analysis To Soil-Landscape Research-Quantifying The Subjective, R. David Hammer, John W. Philpot, Jon M. Maatta

Conference on Applied Statistics in Agriculture

Principal component analysis is a multivariate statistical procedure that can be used to identify factors (correlated subsets of variables) in large data sets. This statistical method appears useful for scientists investigating soil processes, but it has received little attention. Reported applications of principal component analysis share a common fault--subjective, user-specified analytical options apparently are not recognized, for they are not discussed. Reported data sets are often small, have low observations-per-variable ratios, and lack tests of robustness. A large soil data set is used to demonstrate systematic procedures for an optimum rotated principal component solution. This solution retained 21 variables aligned …

Statistical Analysis Of Spectrophotometric Assays In The Presence Of Interference, Edward Gbur, Patti Landers, Roy Sharp

Conference on Applied Statistics in Agriculture

Indirect measurement of the amount of a specified component in a sample of a chemical compound can be accomplished by spectrophotometry. The underlying principle is Beer's Law, which states that, in a pure system, the amount of light absorbed by a chemical bond is linearly related to its concentration. In some mixtures it may not be possible to find a wavelength at which only the bond of interest absorbs light. Hence, the absorbance is composed of contributions from the bond of interest and one or more other (nuisance) bonds. Chemists refer to this situation as interference. In this paper we …

Estimating Mixture Fraction And Map Distance In A Mixed F2, Bc1 Population, Dennis L. Clason, N. Scott Urquhart, Joe Corgan, Catherine M. Cryder

Conference on Applied Statistics in Agriculture

An F₁ interspecific hybrid onion (Allium cepa x A. fistulosum) was backcrossed to the A. cepa parent line under field conditions. The progeny of this cross were shown by electrophoretic protein analysis to be a mixture of BC₁ (the desired backcross) and F₂ (A. cepa x A. fistulosum) x (A. cepa x A. fistulosum) individuals. This mixture of populations among the progeny render the usual X² test for independent segregation of loci invalid. F₂ is used to denote progeny derived from either selfing of the F₁ or from sib-crosses between two F₁ individuals. …

Genotype X Weather Interactions In Grain Yields Of Wheat, Arlin M. Feyerherm, Rollin G. Sears, Gary M. Paulsen

Conference on Applied Statistics in Agriculture

The purpose of this paper is to demonstrate the advantage of using weather elements as covariates in studying yield differentials between varieties of wheat over different climatological regions. Using regression methods, the dependence of varietal yield differences on weather elements was demonstrated with a relatively small sample consisting of yield and weather data over a 3-year period from nine locations in Kansas. For each location, the sample-derived regression equation was used to calculate predicted yield differentials and 95% confidence intervals for the mean (CLM) for each year from 1950 through 1989. The proportion of CLMs that covered positive (or negative) …

Nepotism In Honey Bees, Computer Programs And Scientific Hypotheses, Benjamin P. Oldroyd, Thomas E. Rinderer

Conference on Applied Statistics in Agriculture

Page et al. (1989) attempted to show that bees on queen cells preferentially reared their super sisters as replacement queens rather than half sisters. In support of their contention, they used computer simulation to model the biological system. We argue that the simulation did not accurately reflect the biological system in several important respects. We show that random data will produce the same kinds of statistical significance as the actual data.

Using Response Surface Methodology With A Multivariate Response To Improve The Quality Of A Food Product, George A. Milliken, Tanya W. Maclaurin, Carole S. Setser

Conference on Applied Statistics in Agriculture

Nutrition in health is a major area of focus in our national health priorities as we move into the 21st century. The government, food industry, food scientists, health professionals, and all disciplines that can assist need to work together in the development of "healthful food products" and encourage Americans to make healthful food choices (Drishell, 1990). Experimentation science provides strategies for helping food scientists improve existing food products and develop new ones. This paper describes a process where design of experiments and response surface methodology were utilized in the formulation development to guide product development of a healthful muffin that …

A Simulation Study On The Relationship Between The Abundance And Spatial Distribution Of Insects And Selected Sampling Schemes, J. L. Willers, D. L. Boykin, J. M. Hardin, T. L. Wagner, R. L. Olson, M. R. Williams

Conference on Applied Statistics in Agriculture

During the development of a Bayesian approach to estimate insect population abundance, it was necessary to compare not only the reliability of Bayesian estimates, but to also compare these estimates to those obtained by traditional methods employed by entomologists. To facilitate these comparisons it was necessary to use simulated fields apportioned into quadrats where conditions representative of insect abundance and dispersion are modeled. Thus, a simulation model was developed using SAS to derive example insect populations from which samples could be drawn. The negative binomial distribution was used to simulate the proportion of infested plants (p) with various degrees of …

Composite Sampling Techniques For Determining Pesticide Concentrations, R. S. Parrish, G. O. Ware, C. N. Smith, P. A. Banks

Conference on Applied Statistics in Agriculture

Composite sampling techniques are compared with random sampling methods for determining pesticide concentrations in agricultural fields. Estimates of mean pesticide concentrations and associated standard errors are presented for different experimental conditions. Variance components defined in extended forms of the Brown-Fisher model are estimated. The method of nonlinear least squares was employed to obtain numerical estimates of variance components by equating observed mean squares to expected mean squares for appropriate sampling designs.

A Spatial View Of The Negative Binomial Parameter K When Describing Insect Populations, Linda J. Young, Jerry H. Young

Conference on Applied Statistics in Agriculture

Measures of aggregation as applied to insect populations are reviewed. When these measures indicate strong aggregation, an aggregated spatial pattern is often assumed. The literature noting that the measure of aggregation does not necessarily indicate spatial aggregation, or the lack of it, is reviewed. Field data from four insect species are presented. In each case, the measures of aggregation indicated strong aggregation, but the spatial pattern was not significantly different from random.

Analysis Of Repeated Measures Data, Ramon C. Littell

Conference on Applied Statistics in Agriculture

Data with repeated measures occur frequently in agricultural research. This paper is a brief overview of statistical methods for repeated measures data. Statistical analysis of repeated measures data requires special attention due to the correlation structure, which may render standard analysis of variance techniques invalid. For balanced data, multivariate analysis of variance methods can be employed and adjustments can be applied to univariate methods, as means of accounting for the correlation structure. But these analysis of variance methods do not apply readily with unbalanced data, and they overlook the regression on time. Regression curves for treatment groups can be obtained …

Editor's Preface, Table Of Contents, And List Of Attendees, George A. Milliken

Conference on Applied Statistics in Agriculture

These proceedings contain papers presented at the second annual Kansas State University Conference on Applied Statistics in Agriculture, held in Manhattan, Kansas, April 29 through May 1, 1990.