Biometry
[MAT1375]
[25h + 25h] 4 credits
This course is taught in the 1st semester.
Teacher(s): Nicolas Schtickzelle
Language: French
Level: First
Aims
After completing this course, the student should master the basic methods for the analysis of univariate data, be able to choose the analysis model and method adapted to the design of simple factorial experiments, to analyse and interpret correctly the results of such experiments. He should be able to progress by himself and follow fruitfully advanced lectures on experimental design and data analysis.
Main themes
2) Linear regression: simple, multiple, polynomial.
3) Analysis of variance: one way, two ways, factorial; crossed or nested factors; fixed or random levels; multiple comparisons of means.
The theoretical lectures and practical works shall rely upon real cases. Practical works will be realized using computer software.
Content and teaching methods
2) Analysis of count data: fitting a probability law, two- and three-way contingency tables (homogeneity tests, independence: marginal, partial, conditional).
3) Simple linear regression: model, estimation, inference, validation, prediction.
4) Multiple regression: modelling the effects, model and design, F-tests and AIC criterion; model selection.
5) One-way analysis of variance: principle, models, tests, comparison of means.
6) Multiple-way Anova: concepts of interaction, nesting, fixed or random factor levels, some designs for heterogeneity control.
7) "What next?" Some indications related to the interests of the students (e.g. linear and non linear modelling, non parametric methods, multivariate analysis, geostatistics, Monte-Carlo methods ).
Other information (prerequisite, evaluation (assessment methods), course materials recommended readings, ...)
Course "Mat 1275", Statistics in the natural sciences, or an equivalent course.
EVALUATION
Examination based on solving exercises. Written preparation (in a computer class room), oral presentation.
SUPPORT
Syllabus, booklet of exercises + solutions, overheads, software (SAS). This course is on iCampus, where more information is available and where the yearly calendar of activities is posted.