Extension neglect[a] is a type of cognitive bias which occurs when the sample size is ignored when its determination is relevant.
[1] For instance, when reading an article about a scientific study, extension neglect occurs when the reader ignores the number of people involved in the study (sample size) but still makes inferences about a population based on the sample.
In reality, if the sample size is too small, the results might risk errors in statistical hypothesis testing.
A study based on only a few people may draw invalid conclusions because only one person has exceptionally high or low scores (outlier), and there are not enough people there to correct this via averaging out.
Extension neglect is described as being caused by judgment by prototype, of which the representativeness heuristic is a special case.