Return to the inequality dataset from Lab 2. Fit the interaction model (Gini ~ Polity * log(GDP)) after performing multiple imputation with mice (5 imputations). Compare the results with the complete‑case analysis from Lab 2. Are there meaningful differences? What does this suggest about the missingness mechanism?
Load the AJR replication data (ajrdata.csv). You can find the original article. Replication data are available as well. Replicate the IV analysis from Lab 3 (or a simplified version) using listwise deletion. Then use multiple imputation (mice) to handle missing values in the covariates. Re‑estimate the IV model on each imputed dataset and pool the results using mitools or mice::pool. Compare the coefficients and standard errors. Discuss whether multiple imputation changes the substantive conclusions.
For the AJR data, consider the variable logmort0 (settler mortality). Speculate on whether missingness in this variable is MCAR, MAR, or MNAR. Justify your answer. If it were MNAR, what additional analyses might you attempt?