Heuristics are "rules of thumb", cognitive strategies people use to make selections in the face of data overload. For example, an employer might use the heuristic "long hair means the person is a flake" while making hiring decisions. As in the case of the previous example, heuristics don't always work effectively. Some heuristics lead to systematic errors that can be experimentally isolated -- hence they are labeled biases.
The most common and illustrative example of a systematic bias is to overassign probabilities to conjunctions. Try the following:
Linda is 31, single, outspoken, and very bright. She majored in philosophy in college. As a student, she was deeply concerned with discrimination and other social issues, and participated in anti-nuclear demonstrations. Which statement is more likely?
a. Linda is a bank teller.
b. Linda is a bank teller and active in the feminist movement.
When combined with other options to throw the test-taker off, the majority of people actually pick b, even though the probability of b (a conjunction) is surely lower than the probability of a, which is a superset of b. But our minds automatically work this way. Various heuristics and biases seem to be built into the way our human mind works.
Estimate the product of the series:
9 x 8 x 7 x 6 x 5 x 4 x 3 x 2 x 1 = ?
1 x 2 x 3 x 4 x 5 x 6 x 7 x 8 x 9 = ?
Experimental studies have confirmed that estimates are strongly biased towards the first series. In a study that required participants to give their answers within five seconds, the average estimate for the first series was 4,200, and for the second series, only 500. The real answer is 40,000. Everyone radically underestimated the real answer.
This bias is called anchoring -- fixating on what comes first, and insufficiently adjusting as more data comes in. In a sales context, salespeople will often show a customer a more expensive product, then adjust downwards incrementally. This makes all the products seem cheaper, and is a very effective sales strategy that exploits universal human heuristics and biases.
Bayes' rule has often been cited as a way of making predictions mathematically and normatively, freeing decision-makers from the threat of biased decisions. Unfortunately, applying Bayes' rule in everyday contexts can be difficult for those who are not explicitly trained to do so.