Matching methods is a class of statistical and methodological techniques employed in observational studies to construct comparable groups of treated and control units. By selecting subsets of data or adjusting weights based on measured covariates, these methods aim to reduce confounding bias and strengthen causal inference in non-experimental settings.