Life Sciences Commons^™

Some Statistical Considerations To On-Farm Trials In Kenya, Peter M. Njuho

Conference on Applied Statistics in Agriculture

The issue of design and analysis of on-farm trials is not clearly understood by agricultural researcher in Kenya. Since on-farm trials require participation and co-operation of the farmers who often differ in education level, chances of collecting unreliable and quite variable data are high. This paper highlights the importance of collecting quality data from on-farm trials, and in particular from the researcher designed and farmer managed trial type. Some complexities associated with the implementation, application of basic statistical principles, and analysis of on-farm trials are discussed. Questions that need to be considered priori to the implementation of any trial implementation …

Optimum Design On Step-Stress Life Testing, C. Xiong

Conference on Applied Statistics in Agriculture

This paper presents exact optimum test plans for simple time-step stress models in accelerated life testing. An exponential life distribution with a mean that is a log-linear function of stress, and a cumulative exposure model are assumed. Maximum likelihood methods are used to estimate the parameters of such models. Optimum test plans are obtained by minimizing the mean square error between the maximum likelihood estimate of a certain moment of the lifetime at a design stress and the real moment. The advantage of our optimum test plans is that it does not require large number of items to be tested. …

Modelling The Coefficient Of Variation In Factorial Experiments, Craig A. Wilson, Mark E. Payton

Conference on Applied Statistics in Agriculture

The coefficient of variation (CV) has long been used as a measure of the relative consistency of sample data. However, little attention has been paid to using the CV to make conclusions about the relative consistency of the population(s) from which the data are drawn, particularly when the data are observed in the context of a designed factorial experiment. This research focused on using three approximations to the exact distribution of the sample CV of normally distributed data (McKay's, David's, and Iglewicz and Myers') in the context of the generalized linear model to develop a method for detecting main effects …

A Repeated Measures Analysis Of The Effect Of Vegetative Buffers On Contaminant Runoff From Bermudagrass Turf, Craig A. Wilson, Mark E. Payton, Ellen W. Stevens, Raymond L. Huhnke, Daniel E. Storm, Nick T. Basta, James H. Baird

Conference on Applied Statistics in Agriculture

A repeated measures analysis was conducted on a set of data from a multi-year study to assess the effect of vegetative buffers on the surface runoff of selected herbicides and nutrients. Multiplicative models describing the observed behavior of runoff concentration over time for buffered and non-buffered plots were fitted on a log-transformed scale using linear mixed models with PROC MIXED in PC SAS version 6.11. A spatial power covariance structure was used. Additional models for contaminant mass flow rates were fitted to evaluate the effect of buffers on total runoff mass.

Sampling Plans For Monitoring Rust In Coffee Trees, Raw E. Macchiavelli, Roclo Rodriguez

Conference on Applied Statistics in Agriculture

Rust is an important fungal disease of coffee that causes defoliation of plants, thus affecting production and yield of coffee beans. Management of coffee rust with fungicides should be based on disease incidence at the onset of the epidemic. Previously, sampling has been done in a two-step systematic plan: trees are sampled in a systematic pattern (in the shape of a W covering the field), and then leaves are randomly sampled within each selected tree. Since coffee in Puerto Rico is typically grown in areas with pronounced slopes, these plans require walking diagonally along slopes, which is not feasible for …

Blocking In Partial Diallel Crosses, D. R. Aneja, L . S. Kaushik

Conference on Applied Statistics in Agriculture

Generally the parents are included in the experimental material for combining ability analysis for comparing the performance of crosses with parents and also for calculating heterosis. But unfortunately the parents are ignored for combining ability analysis because of non-availability of analysis procedure. Method of analysis of partial diallel crosses in incomplete blocks for the method -2 of Griffing (1956) has been given.

Construction Of Asymmetrical Response Surface Designs, V. P . Manocha, L. S. Kaushik

Conference on Applied Statistics in Agriculture

The paper proposes several methods for constructing both rotatable and non-rotatable asymmetrical response surface designs. The idea of modified rotatable design is introduced. In most of the experiments conducted by the experimenter it is not necessary that all the factors under study may have equal number of levels The methods proposed will have wider use under these circumstances.

On-Farm Research In A Decentralized Information Model Or Grassroots Statistics, Derrick N. Exner, Jennifer Kendall, Dick Thompson

Conference on Applied Statistics in Agriculture

Practical Farmers of Iowa (PFI) is an organization that seeks to provide interactive methods of relaying information through farmer-to-farmer sharing (farm field days, workshop discussions, networking) and the generation of new information. On-farm research (OFR) is an important information-generating activity of this group. PFI has shown that key to doing research on farms lies in combining practical protocols with the statistician's old familiar friends - replication and randomization.

We provide background on PFI and how PFI cooperators carne to using strip plots and paired comparisons to answer fundamental questions about what to do on their individual farms. We discuss the …

Blocking Factorial Designs In Greenhouse Experiments, Steve Ferris, Steven G. Gilmour

Conference on Applied Statistics in Agriculture

Experiments in greenhouses usually have to be conducted with very limited resources. This makes it particularly important to control the between plot variation by appropriate use of blocking. Many greenhouse experiments are naturally laid out in a pattern that makes a class of designs known as semi-Latin squares useful. Their properties have been studied recently by a number of authors and this work is reviewed. Often, the experimental treatments will have a factorial structure. An example of a 23 structure is used to show how factorial treatments can be assigned to treatment labels to ensure that the appropriate information is …

Daily Solar Radiation Estimated From Tkmpera Ture Records, D. W. Meek

Conference on Applied Statistics in Agriculture

Crop growth models and other environmental analyses require the input of daily global solar radiation values. Unfortunately many locations lack long-term solar radiation data. Most agricultural experiment stations, however, have daily temperature records. Also they are often the locations for which crop growth simulations are conducted. In an unpublished manuscript in the field of agricultural meteorology, researchers wanted to address this need. Specifically they wanted to estimate historical daily global solar radiation using daily air temperature data records by adapting a single published empirical intrinsically nonlinear model, a form of the Weibull curve. In order to help future research in …

Using Confidence Intervals To Obtain A Family Of Estimators Of The Intraclass Correlation Coefficient (Or Heritability), Brent D. Burch, Ian R. Harris

Conference on Applied Statistics in Agriculture

A family of point estimators is presented for the intraclass correlation coefficient (or heritability) in the balanced one-way random effects model. The family is obtained by equating a pivotal quantity to different values of the pivoting distribution, and includes the familiar ML and REML estimators. In terms of mean-squared error, most members of the family of estimators are admissible within the family. A sire model is used to illustrate the estimation of heritability. The authors provide guidance concerning the choice of an individual member of the family for estimation purposes and indicate how the method can be extended to unbalanced …

An Alternative For Mixed Model Analyses Of Large, Messy Data Sets (Mtdfreml), L. D. Van Vleck, R. K. Splan

Conference on Applied Statistics in Agriculture

Portable Fortran based programs (MTDFREML) were developed using a derivative-free algorithm to obtain REML estimates of (co)variance components. Computations are based on Henderson's mixed model equations for multiple-trait models with missing observations on some traits and incorporation of relationships among relatives. Many fixed and random factors are allowed with number of levels dependent on computer memory. Data sets with more than 40,000 genetic effects have been analyzed. Options allow solving MME at convergence. Constraints are automatically imposed. Expectations, standard errors of contrasts of solutions for fixed effects and prediction error variances of solutions for random effects can be obtained. Dimensions …

Statistical Threshold Values For Locating Quantitative Trait Loci, R. W. Doerge

Conference on Applied Statistics in Agriculture

The detection and location of quantitative trait loci (QTL) that control quantitative characters is a problem of great interest to the genetic mapping community. Interval mapping has proved to be a useful tool in locating QTL, but has recently been challenged by faster, more sophisticated regression methods (e.g. .. composite interval mapping). Regardless of the method used to locate QTL. the distribution of the test statistic (LOD score or likelihood ratio test) is unknown. Due to the quantitative trait values following a mixture distribution rather than a single distribution, the asymptotic distribution of the test statistic is not from a …

The Analysis Of Over-Dispersed Count Data From A Single Factor Study, George A. Capuano, Linda J. Young, Nancy L. Campbell

Conference on Applied Statistics in Agriculture

Methods for analyzing over-dispersed count data in a one-way layout were compared using a Monte Carlo study. Several variance stabilizing transformations were examined as alternatives to analyzing the raw data using a general linear model. Additionally, generalized linear models were fit using a log link. For the generalized linear model, three approaches to account for over-dispersion were investigated: (1) a negative binomial distribution with known k, (2) a Poisson distribution with Pearson's X² as an estimate of the scale parameter, and (3) a Poisson distribution with over-dispersion estimated using the deviance. The analysis of the raw data and log …

Risk Factors Associated With Culling Age In Dairy Cattle: Applications Of Frailty Models, Mohamed M. Shoukri, Jan M. Sargeant

Conference on Applied Statistics in Agriculture

Culling decisions for dairy cattle are an important component of dairy herd management. To investigate risk factors for culling, farms (clusters) constitute the sampling units. Therefore, we believe that ages-at-culling may be correlated within farms. The score test on the null hypothesis of no extra-variation in survival data was not supported by age-at-culling data collected from 72 dairy farms from the province of Ontario, Canada. To correct for the intraherd correlation, three modelling approaches were used to fit the data: Population-Averaged (PA) , cluster-specific (CS), and Random Effects Models (RAEM). The modelling approaches are described and compared using the dairy …

Assessing Variability Of Agreement Measures In Remote Sensing Using A Bayesian Approach, William J. Price, Bahman Shafii, Lawrence W. Lass, Donald C. Thill

Conference on Applied Statistics in Agriculture

Remote sensing imagery is a popular accessment tool in agriculture, forestry, and rangeland management. Spectral classification of imagery provides a means of estimating production and identifYing potential problems, such as weed, insect, and disease infestations. Accuracy of classification is traditionally based on ground truthing and summary statistics such as Cohen's Kappa. Variability assessment and comparison of these quantities have been limited to asymptotic procedures relying on large sample sizes and gaussian distributions. However, asymptotic methods fail to take into account the underlying distribution of the classified data and may produce invalid inferential results. Bayesian methodology is introduced to develop probability …

Hurst Phenomenon And Fractal Dimensions In Long-Term Yield Data, Susanne Aref

Conference on Applied Statistics in Agriculture

A fractal dimension may be thought of as a measure of randomness. Fractal dimensions based on semivariograms have been used to determine degree of randomness in yearly crop yields. Through rescaled range analysis Hurst exponents also define fractal dimensions. This method of obtaining fractal dimensions gives more reasonable and sensitive measures than the semivariogram method. To address the inherent randomness due to yearly variations, global trends in yield must be removed before either method is applied. After detrending, a fractal dimension obtained from semivariogram is usually that of a random process. The Hurst method yields an exponent H, which results …

Statistical Analysis Of Field Wheat Varietal Performance Trials, A. M. Feyerherm, R. G. Sears, J. J. Higgins

Conference on Applied Statistics in Agriculture

The purpose of this research was to formulate statistical models and assumptions to apply to the problem of comparing wheat varieties for yielding ability among locations within seasons and over seasons. The methodology could just as well be applied to field testing of other crops for yield or other characteristics of interest (test weight, protein level, etc.)

The methodology approaches the problem of comparing varieties by comparing how well each "measures up" when matched against some common checks. For each variety, the basic data are differences in yield between the variety and the average yield of the checks at different …

Linear And Nonlinear Mixed-Effects Models, Douglas M. Bates, Jose C. Pinheiro

Conference on Applied Statistics in Agriculture

Recent developments in computational methods for maximum likelihood (ML) or restricted maximum likelihood (REML) estimation of parameters in general linear mixed-effects models have made the analysis of data in typical agricultural settings much easier. With software such as SAS PROC MIXED we are able to handle data from random-effects one-way classifications, from blocked designs including incomplete blocked designs, from hierarchical designs such as splitplot designs, and other types of data that may be described as repeated measures or longitudinal data or growth-curve data. It is especially helpful that the new computational methods do not depend on balance in the data …

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

Conference on Applied Statistics in Agriculture

These proceedings contain papers presented in the tenth annual Kansas State University Conference on Applied Statistics in Agriculture, held in Manhattan, Kansas, April 26-28, 1998..