Outline schedule

Draft schedule, 17 Dec 2009.

Day 1 : Monday 18 Jan

• Welcome and introductions.
• What the Boot Camp will cover - an overview of the New Statistics.
• Participants previous knowledge and need for statistics.
   
• What is statistical inference?
   
• Working with samples - sampling error, estimates based on samples, sampling distributions.
• Probability distributions.

• Getting started with R software.

Day 2 : Tuesday 19 Jan

• Review of previous day's concepts.
   
• Ronald Fisher's approach to experimental design and confidence interval estimation.
• Modern alternatives to Null Hypothesis Significance Testing (NHST).

• Working Fisher's example data in R.
• Permutation (or randomisation) tests in R.
   

Day 3 : Wednesday 20 Jan

• Review of previous day's concepts.
   
• Likelihood and Maximum Likelihood Estimation with a simple occupancy example.
• Information Theory (IT) and the use of AIC (Akaike's Information Criterion).
• Ecological models: an example with kill rates analyzed in MS Excel™.

Day 4 : Thursday 21 Jan

• Review of previous day's concepts.
   
• Study design and sample selection, including a range of sampling strategies.
• Using simulated data to check study design (with examples in R)
   
• Recording and managing data.

Day 5 : Friday 22 Jan

• Review of previous day's concepts.
   
• Introduction to Bayesian approaches to analysis.
• The importance of prior information (with examples from medical testing).
• Using "Bayesian posteriors" as the basis for decision making: simple examples where we can estimate occupancy, density or a trend with Bayesian methods.

Optional one-day sessions:

We expect that most people will choose to come to one or maybe two of these. Attendance at the first 5 days is a prerequisite for each of these.

Day 6 : Mon 25 Jan : Occupancy

• The meaning of "occupancy".
• A simple survey of ant occupancy on the lawns at Sama Jaya, with analysis in PRESENCE.
• Incorporating covariates (such as forest type, elevation, observer) into the model; example of camera trap data for golden cats with the PRESENCE package.
• When detection probability depends on animal abundance - the Royle-Nichols model (and analysis in PRESENCE).
• Advanced topics: multiple seasons, multiple detection methods; multiple species with species interactions; spatial autocorrelation; Bayesian analysis methods.

Day 7 : Tues 26 Jan : Distance sampling

• Principles underlying distance sampling.
• Analysis of a simple data set with the DISTANCE package.
• Understanding the output from DISTANCE, in particular goodness-of-fit information.
• Selecting the best detection function.
• Sampling issues: choice of sample frame; layout of transects; stratification.
• Data collection issues: precision of measurements, heaping.

Day 8 : Wed 27 Jan : Mark-recapture methods

• An experiment with mark-recapture where we know the right answer.
• Simple analysis with the MARK package (Khana tiger data).
• Importance of differences in capture probability (ie. models other than M0).
• Estimating the area sampled by the mark-recapture survey: ad hoc solutions and the Royle-Dorazio "spatial mark-recapture" approach.

Day 9 : Thurs 28 Jan : Measuring biodiversity

• What is "biodiversity"?
• Estimating species richness - practical example using EstimateS and data for bats from Loagan Bunut NP; difficulties with estimating richness; alternatives.
• Comparison of richness between sites using rarefaction.
• Biodiversity and evenness indices - discussion and calculation of examples in R.
   
• Differences between sites (beta diversity); Anne Chao's methods of estimation in EstimateS.
• Estimation in R and display using clustering and dendrograms.

Day 10 : Fri 29 Jan : Putting it all together

• Research objectives and identifying variables to measure.
• Research methods - review of techniques available and their assumptions/prerequisites.
• Examples of studies for different purposes.
   
• Course assessment and conclusion.