Comparative Falsification

principle

A cause should explain differences, not merely fit one outcome. If the proposed cause is present where the outcome is absent—or cannot account for variation within the case—it is incomplete or wrong.

An explanation for why Indian women do not work can sound convincing at the national level and still collapse when you travel south. If it cannot explain why female labor participation differs between North and South India, it has not yet explained India.

Make the cause predict the contrast

Comparative falsification turns variation into a test. First identify the proposed cause. Then find places where that cause should produce similar outcomes and places where its presence or strength differs. A useful explanation must survive both comparisons: India against other countries, and one Indian region against another. When the supposed cause remains constant but the outcome changes, something important is missing. The contrast may expose a confounding variable, an interaction with local conditions, or a cause that was never doing the work attributed to it.

Where it shows up

The North–South test

Explanations of Indian women’s labor participation are tested against the internal regional divide. A theory that accounts for the national average but not this split describes the headline without explaining the pattern.

The cross-country test

The same explanations are compared with evidence from other countries. If another country shares the proposed cause but produces a different outcome, the comparison supplies a failure condition rather than another story that can be fitted after the fact.

A failed contrast narrows the theory; it does not finish the investigation

Different outcomes do not automatically prove that the proposed factor has no effect. Its influence may depend on another condition, vary in strength, or be obscured by measurement and competing causes. Comparative falsification is strongest at rejecting an explanation claimed to be sufficient; identifying the actual causal structure requires further evidence.

Write the comparison that could defeat your claim

Tomorrow, take one causal explanation you are using and name two contrasts before defending it: a place where the cause exists but the outcome differs, and an internal split the explanation must account for. If you cannot predict those differences, downgrade the claim from “the cause” to “one possible contributor.”

Episodes that teach this

Why Don’t Indian Women Work and How It Hurts the Economy - FutureIQ 7,497 views
The speakers test explanations by comparing India with other countries and by asking whether each explanation can explain the North-South divide in female labor participation.