An inductive inference that allows to draw a generalised conclusion based on a statistical proposition.
For example:
80 % of all people who come from Scandinavia are blond.
Björn comes from Scandinavia.
Therefore there is an 80 % probability that Björn is blond.
Such a conclusion is valid as long as the statistical probability with which it holds true is not ignored.
It is possible to draw conclusions that are similar in form to a classical syllogism on the basis of statistical statements.
These come with different limitations and possibilities than the classic forms. For example, the classical fallacies of distribution do not necessarily apply, if the distribution is indicated by a statistical dimension.
Conversely, as with all inductive conclusions, the conclusion is only true with a certain probability, which may change at any time due to additional or changed information.
Not least, because inductive rather than deductive reasoning is involved, we need to be aware of possible fallacies of induction that may apply.
The term “ecological fallacy” refers to an inadmissible transfer of a statistical property from a higher aggregation level to a lower one.
The above example avoids this by including the probability with which this is true in the conclusion. If that metric was omitted, it would appear as if every Scandinavian was guaranteed to be blonde, which is of course not a valid conclusion:
80 % of all people who come from Scandinavia are blond.
Björn comes from Scandinavia.
Therefore: Björn is blond.
For more information, see: Ecological fallacy.
In many cases, a confidence interval or other statistical metric can be used instead of percentages. In some contexts, a qualifying adjective such as “many”, “most” or similar may also be appropriate – although these are of course quite imprecise and can easily lead to discussions as to whether a conclusion based on such vague statements is in fact valid.