Using Modern Linear Statistical Models in the Natural Sciences

06-10 février 2023

Stephanie GRAND

Dre Stéphanie Grand, UNIL

Dre Stéphanie Grand, UNIL

Target audience:

This course is designed for researchers wanting to improve their proficiency in the statistical analysis of complex datasets, in which there are many sources of variation potentially influencing the response variable. This is common in fields such as environmental science, soil science, natural resources sciences, ecology, biogeochemistry, hydrology, geomorphology, etc.

The course will be given in English. Pre-requisites include basic classes in descriptive, inferential and multivariate statistics.

Objective:

The objective is to enable participants to apply advanced linear statistical models successfully and rigorously in their research, by:

• - Providing an overview of available linear and non-linear statistical models and their applications;
• - Pointing out practical problems encountered in using these models, and finding solutions;
• - Guiding the interpretation of model output and advanced statistical tests.

This course seeks to blend theory and application to provide a sound understanding of statistical tools appropriate for the analysis of real empirical data. Participants are encouraged to bring their own dataset.

Software:

The course is taught using SAS. SAS is a data management and statistical analysis software known for its wide range of linear modelling capabilities and its computing efficiency. It is a useful addition to the natural scientist’s statistical toolbox. No prior experience with SAS is necessary. Sample scripts will be provided during class.

Concepts taught in class may be approached with other statistical software such as R.

The course (9h15 – 17h) alternates between presentations of statistical concepts, practical examples and work on the participants’ own dataset. Practice sets will be available for participants who have not yet collected their own data.

Module 1: Introduction to SAS 

Module 2: Exploratory data analysis 

Module 3: ANOVA and regression 

Module 4: General linear models 

Module 5: Testing hypotheses 

Module 6: Data structure 

Module 7: Mixed models 

Module 8: The covariance matrix 

Module 9: Generalized models 

Module 10: Non-linear models

Bâtiment Géopolis, Université de Lausanne

1.5

12