Class 1: Single Samples
Data: Excel file Swine Weights.xls
Weight gains of 100 swine during a 20-day period (From Snedecor and Cochran, 1967)
I. Estimating Simple Descriptive Statistics
A. Convert xls file to .txt file for reading by Mathematica
1. Open x
Lesson 16-Correlation Analysis
I. Simple Correlation
A. Correlation vs. Regression
1. Suppose we have the following data on the relationship between femur length
and height from 11 men:
2. Clearly it looks like there is a relationship, but how should we a
Lesson 15-Non-Parametric Methods
I. Why non-parametric methods.
A. Most of the methods we have covered, including regression and ANOVA, assume
that errors are normally distributed. This is true for both least-squares and
likelihood approaches for hypothes
Lesson 14-Bootstrapping and Jackknifing
I. The Bootstrap
A. Purpose
1. The Bootstrap is a method for
a. estimating parameters and their confidence intervals
b. testing for differences in means between two or more
parameters
2. It does not make any assumpt
Lesson 13-Split Plot Designs
I. What is a split-plot design?
A. Split-plot designs have at least two experimental factors.
1. Typically, treatment levels of at least one of these factors cannot be applied to
individuals, but only to groups of individuals.
Lesson 12: Analysis of Covariance
I. Analysis of Covariance: Theory
A. We have so far examined regression analysis, which involves continuous variables,
and ANOVA, which involves class variables. ANCOVA provides an analysis
that includes both continuous v
Lesson 11 Mixed Model Anova
I. What is a mixed model ANOVA?
A. It is an ANOVA that contains both random and fixed effects, typically as crossed
factors.
B. The mechanics of performing the ANOVA are similar to that of the completely fixedeffect ANOVA with
Lesson 10: Factorial ANOVA
I. Types of Factorial ANOVA
A. Randomized block
1. Blocking is a technique that is used to factor out some of the environmental
variance in order to increase the power of the analysis to detect
other effects of interest.
2. For
Lesson 9: Nested ANOVA
I. Theory
A. For this lesson, we will consider Nested ANOVA.
1. Last time, we considered only one factor. Schematically, this may be
represented in the following way:
2. There is one level of groups,and individuals (represented by t
Lesson 8: One-Way ANOVA
I. Basic Theory for Fixed Experimental Effects
A. Imagine we do an experiment in which we establish a number of different treatments,
we subject a number of individuals to each treatment, and then we measure some
variable on those
Lesson 7: Non-linear Regression and Logistic Regression
I. Reasons for doing Non-Linear regression
A. Curve-fitting
B. Hypothesis Testing (e.g. Testing for non-linearity)
II. Polynomial Regression
A. Simplest case: one independent variable.
B. Example: re
Lesson 6: Multiple Regression
I. In this lesson we will cover three topics:
A. Performing regression with more than one independent variable
B. Choosing the "best" regression model
C. Performing contrasts on independent variables
II. The data
A. From Mich
Lesson 5-Simple Regression
I. The Data
A. X: Independent variable-in principle, measured without error
B. Y: Dependent variable-in principle, thought to be determined by independent
variable.
C. Some data: Y: weight loss (mg.) in batches of 25 Tribolium b
Lesson 4: Two-sample comparisons
I. Comparison of means of continuously distributed variable
A. We will use the Bumpas dataset again. This time we will compare the mean lengths
of birds that survived and birds that died.
1. Read in datasets Bumpas1.txt (s
Lesson 3
Single Samples (cont.) and Two-Sample Analysis
I. Estimation of selfing rate
A. Common problem for plant evolutionary biologists
B. Experiment
1. Take 50 plants that are homozygous for allele A1 at a presumably neutral
marker locus, and 50 plants
Class 2: Single Binomial Sample
I. The data
A. This data comes from crosses that one of my graduate students, Tom Chappell did for
his dissertation. He was studying the interaction between morning glories and a rust pathogen.
In one experiment, he crossed
Answer to Lesson 4 Assignment
First, set up the likelihood model.
To do this, recognie there are 4 types of maternal plant: WWAA, WWaa, wwAA, and
wwaa. Let the frequency of A in the outcross pollen that fertilizes WW ovules be p1 , and the
frequency of W