Introduction to Multivariate Ecological Statistics
Please note that this course is no longer running.
You may be interested in our Big Data in Environmental Biology course.
This four-day training workshop in Multivariate Statistics (theory and practice) provides an integrated programme of statistical approaches for research in ecology, field biology, environmental science and management.
The sessions provide an essential learning for those needing a skillset of statistical approaches required for research, or professional and industry-based projects. Delegates will be introduced to skills in experimental design, sampling strategies and data analysis that are essential to the setting up and evaluation of field experiments, landscape scale studies, research into ecosystem services and processes, and assessing the impacts of environmental change or management on biodiversity.
Delegates will be asked to bring their own data sets and there will be opportunities every day, as well as during the final afternoon, to relate their own data to the topic being taught. Focus on the open source statistical package R will ensure that skills are transferable to a wide range of employer and research settings. In addition, the course offers excellent networking opportunities with colleagues and peers from across the environmental sciences.
The course will revise inferential statistics while learning the statistical programming language R, the use of Rstudio, how to import, visualise and analyse data. The core emphasis is on statistical applications rather than the use of R, but Day 1 provides the opportunity for students to become familiar with the software. The remaining three days summarise the role of experimental design and sampling, from small scale factorial to landscape scale correlative studies and introduce more complex analytical approaches used in the exploration, analysis and interpretation of a wide range of experimental designs, data types and spatial/temporal scales. As well as focusing on multivariate statistical approaches, the course introduces more specific analytical methods used in the interpretation of complex community ecology data.
- Update and revise the use of basic statistical techniques for analysing ecological data
- Introduction to R (and Rstudio development environment), learn how to use the statistical programming language R to import, visualise and analyse data
- Revise the theory behind analysing frequencies (Chi-Square and G-tests) and comparing means (Normal distribution, data transformations and t-tests) and learn how to implement it in practice using R
- Understand the use of transformations including log, square root, presence/absence data
- Revise correlations and regressions (coefficient of determination, Pearson's product moment correlation coefficient, simple linear regression, reduced major axis regression) and learn how to implement it in practice using R
- Revise analysis of variance (one-way and two-way ANOVAs) and learn how to implement the tests using R
- Revise non-parametric statistics (Spearman rank correlation, Mann-Whitney U test, Kruskal-Wallis and Wilcoxon's test) and learn how to implement the tests using R
- Understand multivariate ordination based approaches for the analysis of community data; Review the main techniques and applications using R, including PCA, DCA, MDS
- Learn about the potential of other available software including CAP, PRIMER, PC-Ord
- Review approaches to the ordination of community data and underlying explanatory environmental drivers including CANOCO
- Understand replicated randomised block experimental designs and their uses
- Analysis of replicated randomised block experimental designs using R with a series of simple examples with gradual increases in complexity
- Understand strategies for dealing with spatial and temporal data including time series analysis, spatial and temporal GLMs, repeated measures analysis
- Revision of error structure issues in ecological data (binomial, Poisson, Gaussian) and the use of canonical link functions
- Analysis of landscape scale natural gradients in species and their environment
- Understand the effects of variation in sampling effort and species accumulation curves
- Provide an overview of standardised sampling effort, focussing on rarefaction based approaches to account for differences in sampling effort when comparing different sites in experimental studies
- Evaluate the methods used to predict species richness and learn to use species richness estimation software; Introduction to food webs, food web statistics and their interpretation
- Relate student data to each session throughout the course
The core team of three tutors will provide a supportive but rigorous forum in which students can improve their analytical skills as well as address specific questions related to their own research. As well as providing teaching excellence in data analysis, the tutors will use some well-known data sets in the practical sessions e.g. long term monitoring data sets from Wytham Woods, Oxfordshire. Delegates will also be offered the opportunities to bring and analyse their own data during the sessions.
Upon successful attendance of all sessions, delegates receive a University of Oxford Certificate of Attendance on the last afternoon of the course.
Standard rate: £750.00
Standard rate: £750
Early-bird rate: £650 (only 5 places available)
Peter obtained both Bachelor's and Doctoral degrees from Imperial College, London. He has twenty years' experience in applied ecological research, and lectures in population ecology and ecological methods at the University of Oxford. He has recently co-authored the book Ecological Methods with Prof. Sir Richard Southwood, reviewed here, and is a specialist in population dynamics and tropical and temperate crustacean and fish ecology. Peter has worked extensively on the conservation of wetlands.
Toby is a Research Associate in Land Surface Modelling at the Land Surface Processes group, Centre for Ecology & Hydrology (CEH) in Wallingford, UK.
His research area is Environmental Modelling, focusing on hydrological and ecological systems within the terrestrial land surface. This includes ecosystem dynamics (forests, savannas, etc.) and the factors that drive and modify them ranging from soil properties to landscape-scale hydrology to plant ecophysiology. Toby uses large-scale (global or continent-wide simulations using land-surface models: mostly JULES) and ecosystem-level approaches (C budgets, micrometeorology) as well as focusing on species-specific traits as and when appropriate.
Each session is taught through an effective mixture of lecture, group and practical work for which departmental IT facilities will be provided.
Terms and conditions
Terms and conditions for applicants and students on this course
Sources of funding
Information on financial support