Transcript Dan

Overview
Analysis of Research Data
“Fake” Data Generation
Analysis and Comparison
Research Data
• 24 sites
• Trapped over 1 month
• Covariates:
– Temporal
• Moon phase
– Spatial
• Vegetation
• Moisture
• Bare ground
• Statistics needed:
– Capture probability by night
– Effect of moon phase on capture probability
– Effect of spatial covariates on abundance
Data Generation
Sample Poisson for true abundance
Abundance
Moon effect
Spatial effect
Apply capture probability individual by night
Heterogeneity of trap response
TE
Capture history for a site
Goals of Data Generation
• To create a history of deer mice captures
for 24 plots with a known true abundance
known spatial impacts on total abundance
and know effect of temporal
considerations on capture probability
• Apply a treatment effect to half of the sites
– 5%
– 25%
– 50%
Analysis
Capture Histories
Basic Analysis
Hierarchical Analysis
RMark
T Test and CI
CI width, Estimated mean,
Deviance from true mean
Loop
Distribution of statistics for each analysis
Goals of Analysis
• Generate distributions of statistics
generated by each analysis and compare.
• Determine whether there is a difference
between the results of each type of
analysis
• Determine which analysis produces results
closer to the truth