M06
Statistical Modelling
Testing whether patterns observed in mobility and infrastructure data are strong enough to support a claim, and how much distance, cost and infrastructure each contribute.
Data & Inequality
When It's Useful
When a study needs to move from an observed pattern to an estimate of how much each factor actually matters.
STRENGTHS
- Separates the effect of several factors at once.
- Produces results that can be checked and reproduced.
- Works across differently sized datasets.
LIMITATIONS
- Only as good as the variables it's given — it cannot model what wasn't measured.
- Correlation in the model still requires careful interpretation, not automatic causal claims.
- Can obscure individual cases behind an average.
Variables Typically Captured
access scoredistancecostinfrastructure quality indexincome
Process
01
Question
02
Variable selection
03
Model specification
04
Estimation
05
Validation