Propensity Score Matching
Propensity Score Matching - Related Videos
01:30
Marginal Propensity to Consume
0 Views •
2025
The marginal propensity to consume (MPC) describes how much of an additional dollar of disposable income a household is likely to spend rather than save. It provides insight into consumer behavior and is a foundational component in the analysis of fiscal policy effectiveness and national income determination.Concept and MeasurementMPC is measured as the ratio of the change in consumption (ΔC) to the change in disposable income (ΔY), expressed as:MPC = ΔC / ΔYFor example, if an individual's...
06:55
Inverse Probability of Treatment Weighting (Propensity Score) using the Military Health System Data Repository and National Death Index
0 Views •
Cited by 14 •
2020
When randomized controlled trials are not feasible, a comprehensive health care data source like the Military Health System Data Repository provides an attractive alternative for retrospective analyses. Incorporating mortality data from the national death index and balancing differences between groups using propensity weighting helps reduce biases inherent in retrospective designs.
01:26
The Marginal Propensity to Save
0 Views •
2025
The Marginal Propensity to Save (MPS) describes the proportion of additional disposable income that a household saves rather than spends. It is calculated by dividing the change in savings by the change in disposable income. This ratio helps economists understand individual and aggregate saving behavior and is critical in developing models of income distribution and economic growth.Example of MPS CalculationTo illustrate, imagine that Kevin's disposable income increases by one hundred dollars.
01:17
Sign Test for Matched Pairs
0 Views •
2025
The sign test for matched pairs offers a robust method for comparing two paired samples, often for the effects of an intervention in one of them. This method is very useful in situations where the underlying distribution of the data is unknown. The test compares two related samples—often pre- and post-treatment measurements on the same subjects—to determine if there are significant differences in their median values. To conduct the sign test, we first calculate the differences in value between...
06:45
Transmembrane Domain Oligomerization Propensity determined by ToxR Assay
0 Views •
Cited by 7 •
2011
An efficient procedure to assess the oligomerization propensity of single-pass transmembrane domains (TMDs) is described. Chimeric proteins consisting of the TMD fused to ToxR are expressed in an E. coli reporter strain. TMD-induced oligomerization causes dimerization of ToxR, activation of transcription and production of the reporter protein, -galactosidase.
FAQs
Covariate balance makes treated and untreated participants more comparable on the observed characteristics used in the analysis. When matching produces closely aligned propensity scores, it improves the basis for comparing outcomes between groups. This matters because differences in outcomes can then be assessed with less influence from measured covariates that were associated with treatment assignment.
The propensity score summarizes each participant’s conditional probability of receiving treatment given the observed covariates. Matching participants with similar scores creates pairs or groups whose treatment probabilities are aligned, even when their individual covariate values differ. This score-based comparison helps researchers organize observational data for treatment-effect estimation.
Randomization assigns treatment through a study design, whereas propensity score matching works with treatment assignments that already occurred in observational data. Matching can improve comparability and approximate a randomized comparison, but it does not reproduce random assignment completely. Its conclusions remain vulnerable to confounding from factors that were not measured.
Matching addresses differences associated with the observed covariates included in the propensity score, not characteristics absent from the data or analysis. If an unmeasured factor influences both treatment receipt and the outcome, treated and untreated participants may remain systematically different despite similar scores. Consequently, matching strengthens causal assessment without eliminating every source of bias.
Researchers first estimate each participant’s probability of receiving treatment from observed covariates. They then identify treated and untreated participants with closely aligned scores and use those matched comparisons to examine treatment effects. The workflow is intended to improve covariate balance before interpreting outcome differences, while recognizing that unmeasured confounding can still affect the estimate.
This approach is useful when researchers need to study treatment effects but cannot obtain randomized assignments. Its applications include medical, policy, and social science data, where observational comparisons may otherwise reflect differences between those who received treatment and those who did not. Matching provides a more comparable basis for evaluating causal relationships in these settings.