PBR-34803 Experimental Design and Data Analysis of Breeding Trials (DL)

Course

Credits 3.00

Language of instruction:

English

Assumed knowledge on:

MAT-25303 Advanced Statistics (DL)

Continuation courses:

PBR35803 Design of Breeding Programs.

Contents:

Note: This course has a maximum number of participants. The deadline for registration is one week earlier than usual. See Academic Year.(http://www.wageningenur.nl/en/Education-Programmes/Current-Students/Agenda-Calendar-Academic-Year.htm) -> Registration for Courses.
Note: The period mentioned below is the period in which this course starts. For the exact academic weeks see the courseplanning on www.wur.eu/schedule.

In this course, students are taught principles of experimental design of trials and statistical analysis of trial data with a special emphasis to linear and generalized linear methods, mixed models, analysis of multi-environment trials using different statistical methods

Learning outcomes:

After successful completion of this course students are expected to be able to:
- comprehend statistical principles underlying experimental designs for breeding trials with respect to randomization, replication (including types of replicates and pseudo-replication), blocking, experimental units, the use of controls, orthogonality, balance and efficiency, power;
- comprehend the connections between these design principles and the models and model assumptions underlying statistical analyses, most importantly linear regression and analysis of variance (distributional assumptions, independence, equal variance; additivity or linearity of effects, single or multiple random error terms);
- apply these concepts when designing an experiment;
- explain, distinguish and characterize the following experimental designs: completely randomized design (CRD), randomized complete block design (RCB), incomplete block designs (including resolvable designs: lattice designs and alpha designs, row-column designs) and split-plot designs;
- understand effects of missing values, data errors, outliers, uneven replication, confounding of effects, and violations of distributional assumptions and assumptions of equal variance and independence;
- understand when generalized linear models (GLM) are more appropriate for data analysis than linear regression or Anova;
- understand when linear mixed models (LMM) are more appropriate for data analysis than linear regression or Anova;
- be able to perform different analyses using GLMs: logistic regression for binary data, threshold models for multinomial or ordinal data, loglinear regression for counts; - understand how and why distribution and link functions need to be specified in GLMs;
- understand the difference between fixed terms and random terms in a mixed model analysis, both conceptually and in applications;
- specify a linear mixed model in fixed and random terms for a data analysis with unbalanced designs;
- specify a linear mixed model in fixed and random terms for a data analysis with dependent observations;
- comprehend and apply linear mixed models in different contexts: estimation of variance components (e.g. for heritability estimation), or quantify the relative importance of environmental and genetic contributions to the variation in multi-environment trials; analysis of split-plot trials;
- use a linear mixed model for the estimation of variance components;
- explain genotype by environment interaction as a concept in multi-environment trials in plant breeding and in statistical terms;
- quantify, test and characterize genotype-by-environment interaction using different evaluation methods: analysis of variance, mixed models, Finlay-Wilkinson regression, AMMI and GGE biplot;
- comprehend and discuss the concepts of stability, adaptability and (wide/specific) adaptation in plant breeding in the context of Finlay-Wilkinson regression;
- estimate heritability of traits from estimates of variance components obtained from Anova and mixed models in genotype trials.

Activities:

Knowledge clips, individual and group exercises, E-learning modules.

Examination:

The exam is an online remotely proctored exam, where the student should provide a suitable computer and room.

Literature:

Available through the course website.