A hasty generalization error or overgeneralization is a statement based on insufficient evidence. Instead of looking at examples and evidence that fit much more closely with a typical or average situation, you draw a conclusion on a large population using a small, unrepresentative sample.
Cause
A hasty generalization error occurs when people draw a conclusion from a sample that is too small or made up of too few cases.
When we try to understand and come up with a general rule for a situation or problem, the examples we use should be typical of the current situation. If we only consider exceptional cases or only a few examples of a certain phenomenon, we commit an error of hasty generalization. In other words, we jump to conclusions.
An argument based on a hasty generalization moves from particular statements to a general statement. However, deducing a conclusion about an entire class of things from insufficient knowledge about some of its members is a logical leap.
A hasty generalization generally follows this pattern:
We take a small sample from a population, the sample usually being our own experiences.
We draw a conclusion based on this small sample.
We extrapolate our conclusion to the population.
In other words, “if it’s true in this case, then it’s true in all cases.”
How to avoid excessive generalization?
In statistics, the error of hasty generalization is often the result sampling bias (i.e. when using a sample that does not represent the entire population). This can be accidental or intentional, as in the case of misleading statistics. Be careful to make representative sample summaries (law of large numbers or Gibbs sampling for example).
EN 
FR