This resource contains the complete material from the workshop “Revisiting Surveys”, designed to shift the analytical paradigm from the independence of observations toward structural interdependence.
Content Summary
The course covers 5 main modules, available on the project website:
- The Survey as a Complex System: Critique of the atomistic view and foundations of latent social structure (McPherson, Smith-Lovin).
- The Homophilic Force: Empirical estimation of demographic “weights” on the probability of social contact using personal network data (GSS 2004) and Conditional Logit models.
- Social Context Imputation: Methodology to reconstruct latent networks in traditional surveys (ANES 2016) using social distance matrices (Blau Space) and spatial regression models.
- Ising & Network Psychometrics: Application of physics models (Ising) to understand belief systems and political attitudes through regularized partial correlation networks (eLasso).
- Context Challenge: Practical exercises to test new variables and discover effects of “tribalism” vs. social immunity.
Technical Resources
The material is delivered as an integrated Quarto website including:
- Reproducible R scripts for cleaning and analysis.
- Processed Datasets (GSS 2004, ANES 2016).
- Code templates for student use.
- Controlled environment for reproducibility.
