Data 101

Chi-Squared Tests and ANOVA

This notebook builds on the logistic regression model from Week 10 by returning to a continuous outcome and applying tools from this week's lab: model comparison (AIC, BIC, ANOVA) and multicollinearity diagnostics (VIF). The response variable is data_products_score, which measures the quality and availability of statistical outputs produced by a country. Two sub-indicators are used as predictors: data_use_score and data_services_score. Both reflect distinct but related dimensions of statistical capacity, making them natural candidates for a multiple linear regression model. To simplify the analysis, only the 2023 cross-sectional snapshot is used. This avoids the repeated-measures complexity that would arise from using multiple years and ensures the independence assumption for linear regression is more plausible.

Model Selection

Model selection