EXPERIMENTAL DESIGN and STATISTICS WORKSHOP
Experimental Design: Saturday 4 March 2017
Statistics: Thursday 6 April and Friday 7 April 2017
Presenter: Gordana Popovic
Gordana is an ecological statistician. Her research focuses on modelling communities of species, to understand how species interact with one another and the environment. She spends much of her time coordinating short courses and providing statistical support to ecologists at the University of New South Wales, in her role as the statistical consultant for the school of Biological, Earth and Environmental Sciences. In her free time she can be found outside in the sun.
These two workshops are designed to run in conjunction with each other, however you may attend just one if you prefer.
This course will examine the underlying principles of experimental design with a focus on data collection and analyses to explore questions and hypotheses in research.
Introduction to experimental design- Asking well-defined research questions and understanding biologically meaningful changes
Properties of the data collected – Variability of data in space and time affects experimental design and analysis
Randomisation – How to collect data for valid inference
Control – Importance of comparing to a control group
Sample size – Determining whether there is enough data to answer the question, is our sampling scheme concentrating data collection in the right places
Pilot study – Do you need to test your methods in a pilot study
PRACTICAL COMPONENT (using free online tools and Excel)
How to randomise – Create a sampling scheme
Simple power analysis – How much data do we need to answer the question
This course will introduce participants to the R language and RStudio environment. On the first day we start with the basics, how to open R and import data, basic data handling and plotting. We then delve into manipulating datasets to convert between formats (long and wide), subset data and calculate summary statistics (to make tables for reports and manuscripts). We then learn how to quickly create publication ready plots for all data types in R. The second day will showcase R’s statistical capabilities. We will revise some basics, like data types, p-vaues, confidence intervals, and tests like the t-test and chi-square test. We then go on to introduce linear models. These form the basis of all statistical modelling and can be extended to cover discrete data types (generalised linear models) and more complex experimental designs (mixed models) as well as multivariate models for multi species data. We will introduce the lm() function in R, and focus on how to check model assumptions, test a range of hypotheses and understand model outputs.
Bronte Room, Hurstville Training Rooms, 124 Forest Road, Hurstville
The Experimental Design Workshop will be held from 9 am – 5 pm Saturday 4 March.
The Statistics Workshop will be held from 9 am – 5 pm Thursday 6 April and Friday 7 April.
The cost includes workshop attendance, morning tea and afternoon tea (please bring or buy your own lunch).
|Statistics (2 days)||$400||$600|
|Experimental Design and Statistics (3 days)||$550||$800|
What to bring
- Paper and pen for taking notes
- Laptop with web access and Excel (The room will have ample powerpoints and free WiFi)