The following topics will be addressed in the course: linear regression and multiple linear regression, including model formulation, meaning of model parameters, checking model assumptions and prediction; data transformation; experimental design, including completely randomized design, block design and factorial design, and calculating the required sample size to obtain a certain precision; analysis of variance and pair-wise testing; non-parametric tests, including Wilcoxon (Mann-Whitney), Spearman rank correlation, and Kruskal-Wallis; proportion analysis for one population, test for difference between two proportions, and the binomial distribution; contingency tables and the chi-squared tests for goodness of fit, for independence and for homogeneity; multiple linear regression model comparison; experimental design involving factorial design in blocks; selection of variables (quantitative and/or qualitative) to find the optimal linear regression model, including checking assumptions; repeated measurements; and calibration, validation and cross-validation.
These methods are relevant for further data analysis in the biosystems engineering domain. The theory of the course will be supported by practicals in which relevant data sets from the biosystems engineering domain will be analyzed.
This course is tailormade for Bachelor Biosystems Engineering and part of the course is in Dutch. Students of other programs, please follow MAT-20306 Advanced Statistics which covers nearly the same statistics but with more general examples.