Methods in Experimental Ecology II

Methods in Experimental Ecology II
Why?
Pseudo-replication
Unnecessary complicated analysis
Weak inference
Confounding variables
Methods in Experimental Ecology I
Experimental Design and Pseud-replication
Probability Distributions
Data with normal errors
Summary statistics
Confidence Intervals
Regression
Model Selection approaches
Mathematics
Dept.
Computer Science
Dept.
Statistics Dept.
Methods in Experimental
Ecology II
Bayesian and Frequentist approaches
Complex data
General Linear Models
Mix-models
Non-linear Models
Data modeling
Approach
• Understanding the attributes of the data using
generated data
• Implementation of analysis
• Identification of model limitations
• Understanding differences between
approaches
• Application to case studies
Bayesian Analysis of averages
0
100
200
Frequency
300
400
500
Uniformative prior for the mean
-4000
-2000
0
2000
4000
rnorm(10000, 0, sqrt(1/(1e-06)))
300
200
100
0
Frequency
400
500
Informative prior for the mean
0
20
40
rnorm(10000, m, sqrt(1/v))
60
80
20
0
10
Presence
30
40
Analysis of count data
0
10
20
Salinity(ppt)
Gregory Territo (Parkinson’s Lab)
30
40
Analysis of
categorical data
Wildebeest carcasses from the
Serengeti (Sinclair and Arcese
1995)
Females
Marrow
Death
OG
SWF
TG
Total
PRED
32
26
8
66
NPRED
26
6
16
48
Total
58
32
24
114
Males
Marrow
Death
OG
SWF
TG
Total
PRED
43
14
10
67
NPRED
12
7
26
45
Total
55
21
36
112
ˆ
XY ( k )
n11k n22k

n12k n21k
R
• Students can use any platform they feel
comfortable
• However, I will teach using R
because
•
•
•
•
It is free
It has reasonable documentation in the web
It is flexible
It has many applications for biological purposes