Translating social scientific theories into models. Analyzing models using R and lavaan.|
Relation between assessment and objective
The test(s) consist(s) of three parts. One part consists of multiple research questions for which the student has to choose and conduct the correct analysis. Another part requires the student to make correct interpretations of the output of analyses (e.g., output from R and lavaan). Finally, one part is a multiple choice (TRUE or FALSE) test of general knowledge about statistical modeling.
Note: This course can also be taken as part of the honours programme (and then includes additional work and guidance). Contact the course coordinator if you are an honours student.
Note: Students who cannot comply with the entrance requirements mentioned below will be asked to provide further information on their eligibility. The course coordinator will decide on their eligibility.
Assumed knowledge: Anova, regression and correlation. Participants should be familiar with a statistical program, such as SPSS or R or STATA.
Statistics are a tool test whether a theory can be rejected or not. However, social scientific theories are often more complex than the basic relationships that can be tested in SPSS. This course introduces Structural Equation Modeling: a flexible, intuitive technique that will enable you to represent entire theories and their assumptions, and test them on empirical data. Structural Equation Modeling combines factor analysis – tapping into theoretical concepts based on multiple measured indicators – with multiple regression models. It is used to examine whether theoretical constructs are adequately measured, and to test complex theories. We will discuss, among others, the following topics: 1) How can I test whether questions measure what they intend to measure? 2) How to test complicated models (mediation and path models)? 3) Do theoretical models differ across populations or sub-groups in the population?
You will cycle through all phases of social scientific research: Translating a social scientific theory into statistical models, and analyzing those models based on empirical data (provided at the beginning of the course). Finally, you will learn to interpret and report your results.
Analyses are conducted in the statistical software R, and the structural equation modeling package lavaan. No prior knowledge of R is required; this course can serve as a basic introduction to R. We focus on the specific techniques covered in this course. R is a free, open source program, which can be used just as easily for basic t-tests and correlations, as for cutting-edge analyses such as Structural Equation Modeling or machine learning.