Program for the FA1103 Training School
“Experimental Design and Statistical Data Analysis: Applications in R for endophyte-related data “
Évora, Portugal. 2014
Program for the FA1103 Training School (pdf)
Day1
Refresh basic concepts and offer various tools available for the display and optimal extraction of information from data
Random variables: Definition and types. Population and sample. Sampling distributions. Estimation: Methods, point estimation and confidence intervals. Estimation of means and proportions. Sample size for estimating parameters. Hypothesis testing. Type I and II errors: Significance level and power. Tests for means and proportions. Sample size for testing. Exploratory data analysis: graphing tools.
Day 2
Linear regression analysis and Generalized Linear models
The linear regression model, the log-linear model, the gamma model, the log-poisson model and the logit model.
Day 3
How to collect meaningful data - General considerations
Basic purposes of experimentation; Basic requirements; The experimental process; Definition of some key terms
Designing experiments
Replication; Randomization; Blocking
Choosing the factors and their levels:
Factorial experiments; Choosing the levels of a factor; Sequences of experiments.
Day 4
Towards testing more than two means - Analysis of Variance
The one, two and n way model, with and without interactions; models with nested factors. Concept, notation and assumptions. Estimation of effects. Diagnostic checking.
Multiple comparisons procedures for comparing all treatments.
Non-parametric tests: one- and two-sample (for both independent and paired samples) non parametric tests; non-parametric versions of ANOVA
Day 5
Classical designs.
Completely Randomized Design. Randomized Complete Block Design . Split-plot designs and alternatives. Latin squares designs.