Review:

Propensity Score Matching

overall review score: 4.2
score is between 0 and 5
Propensity-score-matching (PSM) is a statistical technique used in observational studies to estimate the causal effect of a treatment or intervention by accounting for covariates that predict receiving the treatment. It involves calculating a propensity score— the probability of treatment assignment based on observed characteristics— and then matching treated and untreated subjects with similar scores to reduce selection bias.

Key Features

  • Estimates causal effects in non-randomized studies
  • Uses propensity scores to balance covariates between groups
  • Facilitates matching, stratification, or weighting methods
  • Reduces confounding bias in observational data analyses
  • Relies on careful model specification for accurate scoring

Pros

  • Provides a robust method to control for confounding variables in observational data
  • Enhances the validity of causal inferences where randomized controlled trials are infeasible
  • Flexible in application—matching, stratification, and weighting options available
  • Widely used and supported by extensive methodological research

Cons

  • Dependent on the quality and comprehensiveness of observed covariates; unmeasured confounders can bias results
  • Requires correct model specification for propensity scores to be effective
  • Can discard data if suitable matches are not found, potentially reducing statistical power
  • Computationally intensive with large datasets or complex models

External Links

Related Items

Last updated: Thu, May 7, 2026, 02:50:55 PM UTC