Hidden Common Cause

If two things are shown to be (strongly) correlated, it is tempting to assume that one causes the other (see PostHocErgoPropterHoc and CorrelationImpliesCausation). But often there simply is a HiddenCommonCause for both. This will invariably cause a correlation between both.

Not to be confused with SpuriousCorrelation? (if you have a sufficiently large number of studies, statistics says, that a few will find correlations, that are - well - random).

Example:

Without further information one is tempted to read it as: But the truth may or may not be: Other examples: If you think, that these are bad examples or SpuriousCorrelation?, don't be fooled. These is a HiddenCommonCause and it is revealed below the spoiler.

I would say these are better examples because the common cause is less obvious. It depends on what you're trying to show I suppose. If the answer is too obvious people will simply reject it as unrealistic, and that they claim wouldn't be fooled by "real" examples.

And the answers are at the bottom of this page ...

See FallaciousArgument

Spoilers follow:

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.


EditText of this page (last edited March 13, 2010) or FindPage with title or text search