Propensity Score Matching
Do certified buildings actually save energy — or do they just happen to be newer and larger? PS-Matching isolates the true causal effect by pairing certified and non-certified buildings with similar observable characteristics. Adjust the caliper and watch how the matched sample changes.
Maximum distance allowed between matched pairs. Wider = more matches, less precise.
Each certified building is matched to its nearest non-certified neighbour within the caliper.
Matched pairs
—
Unmatched treated
—
Est. treatment effect
—
Adjust the caliper width, then click Run Matching to observe the pairing process.