Experimental Design and Data Analysis

URL study guide

https://studiegids.vu.nl/en/courses/2025-2026/X_405078

Course Objective

In this course, students will become familiar with the most commonly used experimental designs, regression models, some nonparametric tests and bootstrap methods. By the end of the course, students will be able to:Identify and describe some of the most widely used statistical models,Understand the basic theory underlying these models,Apply these models appropriately in relevant contexts.

Course Content

The EDDA course introduces a range of widely used statistical models and techniques, including paired and independent t-tests, non-parametric rank-based tests, bootstrap and permutation methods, analysis of variance (ANOVA), ANCOVA, lasso, linear regression models, generalized linear and nonlinear models, and survival analysis. Students will learn how to choose appropriate models for various scenarios, estimate model parameters using available data, and validate model fit. For regression models, proper study design is essential to draw sound conclusions; the course covers well-known experimental designs such as completely randomized and randomized block designs, along with their associated ANOVA techniques. A portion of the course focuses on non-parametric and resampling methods, such as the Wilcoxon test (one
- and two-sample), Kolmogorov–Smirnov test, rank correlation tests, and both permutation and bootstrap tests. The methods will be illustrated using practical examples, with an emphasis on real datasets. The statistical software R will be used for data analysis throughout the course.

Teaching Methods

Lectures, practical sessions.

Method of Assessment

Course assessment consists of assignments (33.3%) carried out in teamwork during the course and an individual digital test (66.7%) at the end. To pass the course, both components must meet minimum required thresholds. Re-examination policy: if the final result is insufficient, only the individual test may be retaken.

Literature

will be communicated on Canvas.

Target Audience

MSc AI, CS

Additional Information

Assignments are to be made using the statistical package R (http://www.r-project.org).

Recommended background knowledge

A basic course in statistics of the same level as Statistical Methods (X_401020) or Statistical Methods for AI (XB_0080) is mandatory. All the material and its prerequisites (basic mathematical and programming skills) for these courses will be assumed known in the lectures/assignments/exam of the course EDDA.A good background in probability, at least of the same level as in Statistical Methods (X_401020) or Statistical Methods for AI (XB_0080).Basic knowledge of the statistical software R and its application to data analysis.

Explanation Canvas

All the information about the course will be available on Canvas.