Research that shows that people infer causation from correlation

Question

It is well known and frquently quoted that "correlation does not imply causation".

Nonetheless, I've scoured the wikipedia article "Correlation does not imply causation" and the relevant keywords in google scholar, and could not find any research that clearly shows that people use correlation to infer causation. There are some examples from popular newspapers and mentions of "cum hoc ergo propter hoc" as a long known fallacy, but no empirical research.

Is there any research that shows that people have a "correlation implies causation" fallacy?

Answer 1

A much-cited reference on the statistical backgrounds is this one:

• Simon (1954), Spurious Correlation: A Causal Interpretation, J Am Stat Assoc; 49(267)

A more recent, open access but applied research paper on the topic is:

• Parise et al. (2013), When Correlation Implies Causation in Multisensory Integration, Curr Biol; 22(1): 46–9