psy112 - Research methods II - Statistical Learning

Enrolment in Master's programme Neurocognitive Psychology.

Goals of module:
Building upon the basic knowledge in multivariate statistical modeling covered in psy111, after completion of this module students will know how to deal with big data to address empirical questions in neurocognitive psychology. They will be able to solve prediction and classification problems to the realm of basic and applied statistical/machine learning purposes. Furthermore, students will understand the specifics of applied research and the statistical modeling of noisy, longitudinal data.

Competencies:
++ interdisciplinary kowledge & thinking
++ statistics & scientific programming
++ data presentation & discussion
+ independent research
+ scientific literature
++ ethics / good scientific practice / professional behavior
++ critical & analytical thinking
++ scientific communication skills
+ group work

Part 1: Statistical / machine learning methods

• Supervised and unsupervised statistical learning and prediction
• Resampling methods
• Regularized regression
• Linear and quadatic discriminant analysis
• Naive Bayes algorithm
• Tree-based methods
• Support vector machines
• The basics of neural networks
• Principal component regression
• Clustering methods
   

Part 2: Statistical / machine learning methods with R (voluntary hands-on seminar)

• Data examples and applications of the basic machine learning methods covered in the lecture 
   

Part 3: Evaluation research (seminar with theory and practice)

• Paradigms and methods in applied evaluation research (quantitative, mixed-methods)
• Types of studies and designs in evaluation research (experimental, quasi-experimental, (multiple) time series, etc.)
• Multivariate statistical modeling of change over time and group differences in change
• Specific statistical tools for sampling and matching (e.g., Propensity score matching)
• Basics of causality theory and the estimation of average and conditional effects in EffectLiteR
• Research synthesis and meta-analysis

end of summer term

The module will be tested with an oral exam (25 min).

Required active participation for gaining credits:
attendance of at least 70% in the mandatory seminar within one semester (will be checked in StudIP)