Survey Data Analysis in R and Stata

In the first part of the course, you will learn about how survey data is different from other data. We will discuss how you can use survey constructs in your research and what steps are necessary to make sure you can trust your survey responses and measuredconstructs. Using example data, we will go through every step from data preparation to factor analysis andconstruct validity analysis.

In the second part of this course, you will learn about the various tools that you can use to analyze (causal) relationships between survey responses and measured constructs to answer your research questions. We will discuss univariate analyses, structural equation models (SEM) for measurement and causal modelling in Stata and potential pathways for limited sample sizes, i.e. partial least squares (PLS) models. We will deal with the analysis of clusters and latent profiles that can befound in survey data, how to interpret the results andhow to use them in further analyses. We will introduce tools relying on Boolean algebra to solve complex multiplicative relationships. For example, you will learn the basics about qualitative comparative analysis (QCA) and necessary condition analysis (NCA), how these approaches are different from multivariate regression analyses and how to apply them to your (survey) data.

• We will use a slide deck to discuss the main aspects of doing survey research.
• We will go through the code and commands to understand how to use them.
• You will apply what you have learned to solve exercises relying on example data sets.

• After the course, you will write a short proposal (app. 5 pages) containing the theoretical framework, research design (constructs, measurement model and method) and initial results (if already possible) of your research paper.
• Your proposal will be evaluated based on the applicability of research design and analyses for your conceptual framework and the correct interpretation of your research findings (if already possible).