BIOL6052 Advanced Quantitative Methods
Evidence-based conservation of wildlife requires quantitative analyses of field data typically collected under imperfectly controlled conditions and across heterogeneous habitats. This module will develop generic skills in (1) the design of data collection protocols, particularly for field experiments and observational studies, and (2) the testing of hypotheses with statistical models. Although basic statistical awareness will be assumed to undergraduate level, the first third of the course will review core principles of experimental design and analysis that underpin all quantitative methods. The second third of the course will involve developing and applying statistical models for increasingly complex datasets of types commonly encountered in fieldwork, using the freeware environment R (www.r-project.org), the most powerful statistical tool available. Ecological questions commonly seek to explain response variables in terms of predictor variables that co-vary with each other or with nuisance variables, or cannot be measured in balanced designs. The final third of the course will treat these issues in multifactorial and multivariate analysis of ecological data.
Aims and Objectives
1) To teach the principles of experimental design. 2) To develop analytical skills to a level sufficient to understand the principles of field-based data collection and the statistical modelling thereof. 3) To introduce and use the R environment for statistical analysis, graphical figure generation ad result presentation.
Having successfully completed this module you will be able to:
- Design a logical and suitable data collection protocol for statistical analysis of a test hypothesis;
- Identify statistical procedures appropriate to different types of hypotheses and data;
- Interpret the results of statistical tests on given data sets, and use that interpretation to justify the conclusions you draw;
- Compute and present statistical analyses and publication-quality figures using the R environment;
- Independently use freely available software, and know where to search for additional online resources.
The course will start by introducing study design. Various types of statistical analysis will be covered, including: Regression, ANOVA (using the General Linear Model), ANCOVA, the use of Linear Mixed Models and Generalized Linear Models. Students will be taught how to analyse data using the models covered using the statistical environment ''R'' in computer lab sessions. The module will also cover the process of grant proposal writing for the final piece of the assessment.
The module is not subject-specific, and all skills developed on it are transferrable across scientific disciplines.
Learning and Teaching
Teaching and learning methods
• Formal lectures will provide the framework of core concepts and issues. • Discussion workshops will be interactive sessions. The first one will address common issues in the logical design of data collection; the second will be a critique of a Case for Support from a grant application, addressing issues of clarity in presenting test hypotheses and statistical models. • Computer labs will introduce and train in the use of statistical packages for hypothesis testing and presentation of results.
|Practical classes and workshops||12|
|Total study time||150|
Resources & Reading list
Doncaster, C.P. & Davey, A.J.H. (2007). Analysis of Variance and Covariance: How to Choose and Construct Models of the Life Sciences.
Beckerman, A.P. & Petchey, O.L (2012). Getting Started with R.
Crawley, M.J (2008). The R Book.
Fox, J. & Weisberg, S. (2011). An R Companion to Applied Regression.