Popular statistical method for matching data may actually increase bias and imbalance

Popular statistical method for matching data may actually increase bias and imbalance
Image: Intellect (generated)

In short: Researchers analyzing propensity score matching (PSM) found that this widely used technique for causal inference often fails to achieve its intended purpose. Instead of balancing datasets, the study observed that PSM can increase imbalance, inefficiency, and bias, sometimes even making data less balanced than the original unmanipulated set.

The researchers found that PSM's flaws arise because it tries to mimic a completely randomized experiment rather than a more efficient, fully blocked design. This approach leaves the method blind to significant imbalances that other matching techniques could resolve.

The study shows that in datasets already approaching complete randomization, PSM can act like random matching, which actually increases imbalance relative to the starting data. While the authors suggest replacing PSM with other matching methods, they note that propensity scores still have other useful applications.

Why it matters: Researchers using propensity score matching should consider alternative matching methods to avoid introducing bias.

Opens the Intellect+ feed with this card first. Swipe up for the next one.

Sources

To cite this card, link to https://intellectplus.co.uk/app/science-research-propensity-score-matching-may-increase-data-imbalance-and-bi.

How we source, check and correct cards: Editorial standards.

General information, not professional advice. Spotted an error? Flag the card or see corrections and complaints.

Intellect Plus is a visual educational discovery platform for exploring fascinating facts and ideas. Brain Food, Not Brain Rot. What is Intellect Plus?

Related

Finding something worth knowing…