Skip to content

Aún no está disponible en español. Ver todo lo disponible en español

Stockertim Dictionary · Technology

Confounding variable

Also called confounder, confounding factor, lurking variable, confounding

In short

A third factor that influences two series at once and makes them appear linked, even though neither one directly affects the other.

  • Intermediate
  • Concept
  • Checked

In detail

A confounding variable sits behind an apparent relationship. Ice cream sales and drowning incidents both rise in summer; the hot weather drives both, and banning ice cream would not make swimming safer. In finance, the economic cycle is a common confounder: in a boom, company profits, house prices, car sales and share prices all tend to rise together, so any two of them will look correlated even if one has no effect on the other.

Confounding is a problem because it leads to wrong conclusions about cause. An analyst might find that firms with more board meetings have lower returns, when in fact struggling firms hold more meetings and also earn less. The meetings do not cause the poor returns; the underlying difficulty causes both. Acting on the false link, such as cutting meetings, would not help.

Researchers deal with confounders in several ways. Randomised experiments assign treatment by chance, so confounders are spread evenly. In observational data, analysts include likely confounders as controls in a regression, compare like with like, such as firms of the same size and sector, or look for natural experiments where something changed for one group and not another. None of these methods can rule out a confounder nobody has thought of, which is why causal claims from market data need caution.

Worked example

Over ten years, a researcher finds that months with higher advertising spending across the economy are also months with higher share prices.

  1. Check a possible confounder: economic growth, which raises firms' marketing budgets and their profits
  2. Compare months with similar growth: the link between advertising and share prices largely disappears

Economic growth explained both series; advertising was not driving the market.

In practice

Central banks, statistics offices and academic researchers control for the business cycle, inflation and sector when testing what drives returns or growth. Policy evaluation guidance in government sets out methods for separating a programme's effect from confounding factors.

Watch out

  • Adding more controls does not remove a confounder that was never measured.
  • A strong correlation can be entirely due to a confounder.

Learn it properly

Official sources

Facts last checked on 4 November 2026.

Education, not advice: the Stockertim Dictionary explains what terms mean. It never recommends buying or selling anything.

English