Statistical Biases

It’s a family of biases that describes the general tendency to reason against the laws and principles of statistics.

Selection Bias

It can influence a research because of a wrong choice of subjects/data: the observation on data is not generalizable to the whole population.


The tendency to think about risks in terms of proportions instead of differences.


The tendency to remain faithful to our idea, instead of stating the correctness and validity of new information and ideas.

Systematic Bias

Bias that is generated within an organization/system, in which systemic and not accidental errors occur.


The tendency to consider more probable single components, rather than something in its entirety.


Tendency to round off numbers, even though we want results to be precise.

Optimism Bias

Propension to consider the chance of a negative event occurring smaller than it really is.


In a statistical study significant data have been observed. They emerged since they were within specific dimensions and a certain number of cases was analyzed.

Illusory Correlation

It’s the tendency to find correlations and cause-effect relationships even where they are not present.

Gambler's Fallacy

An event is considered less probable if a series of these events has occurred previously in a frequent way. It’s the tendency to think that if something happens frequently, in the future it will happen with a lower frequency and vice versa.

Survivorship Bias

Error in the choice of subjects/data because the choice is based on successes and positive results, without considering the negative ones.

Base Rate Fallacy

It’s the propension to not consider general information, giving more relevance to particular or specific cases.

Hot-hand Fallacy

Having accomplished a series of successes, there is the tendency to think that other positive results will be obtained.

Time-saving Bias

Error in estimating how much time is wasted or saved based on the increase or decrease of speed at which you are travelling: going faster, the time saved is overestimated.

Normalcy Bias

Tendency to underestimate danger or emergency signals and to consider not probable the occurrence of problems.

Pessimism Bias

Tendency to overestimate the probability of negative events occurring and consequently underestimate the probability of positive events occurring.

Aggregate Bias

The idea that aggregated data are not applicable to the single patient (in the medical sphere), this can lead to errors in diagnosis.

Fixed-Pie Bias

There is the tendency to consider a person’s interest as completely opposite to those of another person.

Category Size Bias

It’s the tendency to be more affected by the influence of alternatives that are part of bigger categories, because we think that they could easily happen.

Zebra Retreat

It’s the tendency (in the medical sphere) to hesitate in making a diagnosis for a rare disease, even if it’s the most probable one in that case.