Cognitive bias

From supermemo.guru
Jump to navigation Jump to search

This article by Dr Piotr Wozniak is part of SuperMemo Guru series on memory, learning, creativity, and problem solving.

Definition

Cognitive bias is a brain's natural departure from a mathematically objective computation. Instead of relying on objective logic, the brain uses heuristics to maximize gains while minimizing the cost of computation.

Value of cognitive biases

Cognitive bias in psychology is usually used as a pejorative term expressing weaknesses of human mind. Biases are lumped together as an expression of human irrationality or as errors in reasoning. Instead, all cognitive biases are side effects of the powerful conceptual computation. Cognitive biases are a side effect of the brain's superiority over a digital computer (AD 2020).

Cognitive biases are the cost the brain pays for being an intelligent processor

Mechanism

Most cognitive biases stem from the process of generalization in which the brain simplifies models of reality by forgetting factual details that are secondary to the model. Alternatively, the brain may simplify models and the computation by ignoring excess input data. In logic, the process is known as induction. Generalization is often taken to a highly imprecise extreme. Generalization makes it possible for the brain to use simple computation to reason about complex reality.

One of the most important biases is the confirmation bias in which the evidence in favor of a model is easier to assimilate than the evidence against the model. When a model emerges in the conceptualization process, the stability of its constituent connections will gradually increase. This will make it harder and harder to break up the model with new evidence to the contrary. This is why the world of science often awaits funerals before a new paradigm shift may occur.

The existence and importance of the confirmation bias implies that science should favor free evolution and competition of models (see: Value of wrong models). The brain will naturally protect the coherence of its models to facilitate efficient operations of the organism in the environment (see: Brain algorithms protect models of reality).

Confirmation bias serves model protection and favors the evolution of models in collective intelligence

The art of persuasion

The brain's ultimate decision on the veracity of the model will often be seeded early on the flimsiest of evidence. It can then grow in strength via confirmation bias. This is why the art of persuasion should include a gentle introduction of new models to the target audience. Contrary or aggressive introduction may provide wrong seeding of the model due to the phenomenon of reactance (i.e. natural opposition to coercion or aggression). This is why Richard Dawkins will always preach to his faithful choir, anger the opposition, and make few converts. In the end, it does not matter much that his message is verifiably correct.

Aggressive persuasion may seed strong models that will oppose the promoted models

Abstraction of bias list

The list of cognitive biases is endless, but they all can easily be explained by the power of generalization and conceptual computation heuristics. The power of the human brain resides in the ability to build abstract knowledge. In this context, abstract knowledge of the mechanisms of a cognitive bias makes it possible to predict the existence of all imaginable cognitive biases without actually running psychometric experiments. Most importantly, the power of generalization explains why some biases are seemingly contradictory (e.g. worse-than-average effect seemingly contradicts illusory superiority). Contradictory biases may be associated with brain states (e.g. good mood), disposition, or conceptual activations (e.g. in a debate, the biases will serve the defended point of view). Brain's tendency to see things in black-or-white leads to the separation of opposing camps. For example, some people will recall their school days as bad, others as good. Few people will have a good recall of the objective mix of black, white, and gray. The ultimate outcome of long-term generalizations of the past can be pretty chaotic. It may depend on minor details that often swing the balance due to a confirmation bias in an entirely unrelated area (e.g. judgement on schools swung into an extreme by a nudge of the judgement of a single romantic relationship).

It is important to understand the mechanics of a cognitive bias to prevent endless expansion of the bias list. For example, I am tempted to add glorification of schooling to the list even though glorification is derived from generalization (school is good, or school is bad), and survivorship bias (winners tend to praise their school experience). In turn, survivorship bias itself is based on a projection that generalizes the odds of survival and overestimates retrospective odds. The chain of biased links reflects the structure of the concept map that provides the estimates of odds and values.

Cognitive bias weapon

The list of cognitive biases keeps expanding and its popularity keeps increasing. One of the reasons is that biases are used as a debate weapon. Each time a generalization is used in a debate, the author may be attacked with a cognitive bias sword. The more specific the bias, the more powerful the impact of the attack. If the audience has never heard of a given bias, its impact may be doubled. If a bias is labelled with a name that sounds scientific or erudite, the rhetoric effect may be increased further. Finally, if the mere fact of being affected by the bias undermines someone's intellectual standing, the bias sword may have a paralyzing effect. For all of the above reasons, Dunning–Kruger effect will keep growing in power until its invocation becomes so prevalent it will become trite. People will always tend to over- or under-estimate their expertise or cognitive powers. The reason for that dichotomy of estimate is similar to for why some people are pessimistic while others are perpetual optimists. There is no need to gloat over somebody's overestimate. It adds unnecessary noise and emotion to a debate.

School is good

In this book, I list many cognitive biases that make it hard to eradicate the cultural conviction that "schools are good". That conviction was valuable when it was time to popularize education and the claim that "knowledge is power". Today, children at school experience cognitive dissonance. On one hand, they keep hearing good things about school, on the other they see the reality. Now that they can get all their knowledge on their own, we have to replace "school is good" with "knowledge is good".

On my list of school-related biases, in addition to the glorification of schooling, I often mention the old soup problem, or how toxic memories lift the relative importance of affected asemantic material. Proponents of direct instruction often fall victim of the bias that stems from the fact that the more you teach, the more you need to teach. Proponents of grit fail to notice that the harder you learn the more painful the learning, the less you learn, and the harder you need to learn. Educators suffer from a "best student bias". Metaphorically speaking, they believe that to get more Kipchoges, kids should be paced for a 2-hour marathon. Parents believe in the myth that students are lazy. The more they ban videogames, the bigger the attraction, and the stricter the measures they need to employ. This shows that even in a small area of schooling, the list of cognitive biases has no end. This is why it is important to understand the common denominator rooted in conceptual computation. Most of all, we need to remember that cognitive biases are actually an expression of brain power.



For more texts on memory, learning, sleep, creativity, and problem solving, see Super Memory Guru