My Learning NotesBias
Bias is an inaccurate and irrational evaluation or judgment that leads to wrong decisions and therefore behaviours.
It is important to know what are these biases in order to be able to take more rational decisions.
Here below I report a list of biases that I have found on different sources:
| Bias | Description | Example |
|---|---|---|
| Representativeness | the tendency to judge the likelihood basing on the similarity to the typical case do not considering the real likelihood | If I say: Paolo is introvert and likes math. Is more likely he is a engineer or a worker? If one says engineer because he shares some typical characteristics of an engineer is victim of this bias because they are not considering that workers are more than engineer and therefore it is more likely he is a worker. |
| Insensibility to the prior probability | the tendency to do not considering the real likelihood and focusing on some specific information that may not be indicative or relevant | If I say: 90% of people that work in a company are workers and 10% are engineer and Paolo works in this company. Paolo is very precise and likes logic. Is more likely he is a worker or an engineer? If one says engineer because they use specific information like he is precise is victim of this bias because they are not considering the real likelihood. |
| Insensibility to the sample dimension | the tendency to evaluate in the same way likelihood based on small samples (less reliable) and on big samples | Let’s suppose that there are two batches of components. One batch of 100 pcs and one batch of 10 pcs. The defective pieces are typically the 2%. In some cases the defective pieces in the batch are the 10%. Is more frequent to have the 10% defective pieces in the 100pcs batch, in the 10 pcs batch or in both the batches? The right answer is in the 10pcs batch because in the small batch is more likely to have more variability. The casual variation affects more the likelihood with few samples. |
| Wrong evaluation of the chance | the tendency to consider that the sequence of an event follows the likelihood for both small and big populations. | Let’s consider the toss of a coin. If you had a sequence of HEAD-HEAD-HEAD, you may think that, in the next toss, it is more likely to get at least one TAIL to restore the 50-50% likelihood. This is not correct because each coin toss is independent and the likelihood remains 50-50% regardless of the previous sequence. |
| Illusion of validity | the tendency to consider accurate our forecasts also if the statistics show that most of the time we are wrong. | |
| Wrong conception of the regression | the tendency to do not considering that after a maximum or a minimum there is a regression towards the average. | A basket player scores 40 points in a match and his average points per match in the season are 20 points. It may be considered that he is improving, but it is likely that it is only a coincidence and in the next matches he would score near 20 points (regression towards the average). |
| Availability | the tendency to consider more likely an event that we remember easily because it is happened recently or because it was emotionally significative. | Some people consider planes as more dangerous than cars because plane accidents are more widely covered in the media event though car accidents are statistically far more common. |
| Wrong evaluation of the combined and separated events | the tendency to overestimate combined events (events that must happen together to get a result) and to underestimate separated events (events that can happen separately to get a result). |