Applied Statistics II: B

Full course description

Theme 1, Period 4, offered in PSY4163 & PSY4164

Course lecturer: Gerard van Breukelen

Sample size calculation and nested designs: This course provides an introduction to sample size/power calculation for elementary and often encountered research designs in psychology and neuroscience. First, sample size calculation is explained and practiced for comparing two independent samples (e.g. parallel groups or between-subject design) and for comparing two dependent samples (e.g. crossover or within-subject design) on a quantitative dependent variable (outcome). Subsequently, this is extended to a) correlation between two quantitative variables, b) the comparison of two groups on a binary outcome, and c) two-way factorial designs (BS*BS, WS*WS, BS*WS). The opposite effects of a covariate on the sample size needed in randomized and nonrandomized studies are also explained and practiced. Finally, the data analysis and sample size calculation are covered for some popular nested designs, specifically cluster randomized trials and multicenter/multisite trials. Sample size calculations will be done with GPower and possibly some free software for nested designs, and with pencil-and-paper assignments.

Theme 2, Period 4, offered in PSY4163 & PSY4165

Course lecturer: Nick Broers

Structural equation modeling: Structural equation modeling (SEM) is an advanced multivariate method that is gaining importance in psychology but still requires special software (such as Lisrel, EQS, AMOS or Mplus). SEM is introduced in two units, starting with causal modelling and mediation analysis in cross-sectional research and then extending to longitudinal research and latent variables (factors). Special attention is given to identifying models, model equivalence, global and local goodness of fit indices, parsimony, model modification and cross- validation. Some concepts from matrix algebra are needed for SEM, and these will be briefly discussed without going into technical detail.

Theme 3, Period 5, offered in PSY4164 & PSY4165

Course Lecturer: Jan Schepers

Resampling methods in statistics: Many modern statistical analyses make use of resampling methods in applications where theoretical statistics cannot readily provide answers for making statistical inferences from the data at hand. This elective provides an introduction to three important resampling methods, bootstrapping, permutation testing and cross- validation, for obtaining measures of accuracy for parameters of a model or for studying model fit. The methods will be practiced using the software R.