Hindsight error

The highsite error, although originally bearing a different name, became the object of psychological research in the 1970s. However, the phenomenon itself was not new, and by that time there were already many descriptions of this phenomenon in the works of historians, philosophers, physicists, as well as in works of fiction ^[3] . In 1973, Baruch Fishhoff ^[4] attended a seminar in which Paul E. Mil expressed the idea that clinical physicians, as a rule, overestimate their ability to predict the outcome of a medical case, saying that they knew the result in advance. B. Fishhoff, a psychologist by education, saw the possibility of psychological research in explaining these observations ^[5] .

In the early 1970s, the study of heuristics and prejudices was a popular area of ​​psychological knowledge led by Daniel Kahneman and Amos Tversky . Based on the ideas developed by D. Kahneman and A. Tversky on the heuristic of accessibility and the heuristic of representativeness , ^[6] B. Fishhoff, together with Ruth Beit ^[7], organized the first experimental study of the highsite error ^[8] . The study participants were supposed to assess the likelihood of several possible outcomes for the upcoming visit of US President Richard Nixon to Beijing and Moscow . Some time after Nixon returned, the subjects were asked to reproduce the probabilities that they attributed to each possible outcome at the first stage of the study. It turned out that the subjects greatly overestimated the indicated probabilities of events that had already occurred.

In 1975, B. Fishhoff developed yet another method for investigating the highsite error, which was later called the “reptile determinism hypothesis” ^[3] . In the framework of this method, the subjects are given a short story with four possible outcomes, and they also receive information about what the outcome actually happened. Then the subjects are asked to assess the likelihood of each of the possible outcomes ^[3] . Regardless of what the outcome was called by the researchers to occur, the subjects evaluated the likelihood of such an outcome being significantly higher than all other possible outcomes.

Having developed from the heuristics of A. Tversky and D. Kahneman into the hypothesis of reptile determinism, and then into the error of the highsite, this phenomenon has many practical applications today and is still actively studied by psychologists.

According to Hartmouth Blanck ^[9] and his colleagues, all the descriptions of the highsite error existing in the literature can be reduced to three independent processes: the effect of inevitability, the effect of predictability and distortion of memory ^[10] . All three phenomena represent various possible manifestations of a highsite error, which can occur both individually and collectively.

The first phenomenon - the effect of inevitability - reflects a retrospective increase in subjective probability, or perceived inevitability, of one or another outcome. In other words, when a person learns about how the event ended, this outcome begins to seem to him more probable, or inevitable, than before he acquired this knowledge. This phenomenon was called in the works of B. Fishhoff "creeping determinism" ( English creeping determinism ) ^[3] .

The second phenomenon is the effect of predictability : people tend to believe that they knew everything in advance or were able to predict how this or that event will end. B. Fishhoff called this effect “I knew this from the very beginning” ( Eng. I-knew-it-all-along ) ^[11] .

Finally, the third phenomenon is memory distortion . After receiving an answer to a question (for example, “What is the length of the Danube River?”) Or after receiving information about the outcome of an event, the knowledge in the memory is distorted, adapting to the information received about the correct answer ^[3] .

In the literature, the hindsight phenomenon is most often associated with the predictability effect, followed by the inevitability effect and memory distortion processes ^[10] .

H. Blank notes that these phenomena do not reflect the same phenomenon, since, firstly, each of them is based on various psychological processes ^[10] . Thus, the processes of causal attribution , namely, the processes of creating and modifying cause-effect models of an event, underlie the effect of inevitability. In hindsight, people build and add new cause and effect relationships in such a way that the outcome seems predetermined in advance. The predictability effect, in turn, is associated with metacognitive processes, such as a person’s perceptions of the extent to which he was able to predict a certain outcome. For example, drunk drivers can retroactively admit that after six glasses of whiskey the likelihood of an accident in a car accident is very high, but they will also be sure that they themselves were not able, when drunk, to foresee such an opportunity. Finally, memory distortions are controlled by various memory processes, especially such as linking (“fitting”) the memories to the actual outcome of the event and restructuring the original assumptions.

In addition, as H. Blank notes, all three forms of hindsight differ in the functions that they perform ^[10] . The effect of inevitability, given its dependence on causal representations, performs the function of satisfying the need for control . Causal models constructing cause-effect relationships make events more manageable and potentially predictable in the subjective perception of the individual. Moreover, the effect of inevitability can also help to cope with disappointment from undesirable outcomes (“I had no chance”) ^[12] . The predictability effect, in turn, serves the function of self-affirmation (for example, it contributes to the perception of oneself as an informed person, including in the eyes of others) ^[13] . In addition, subjective predictability (or unpredictability) performs the function of self-defense (maintaining a positive personal identity) in the case of a negative outcome, for which a person can be held responsible. For example, owners of stocks that have fallen in price believe that they were not able to predict failure in advance ^[14] . Memory distortions, according to some authors, can be considered as a by-product of the processes of updating knowledge and, thus, can indirectly affect the function of updating knowledge ^[15] .

The role of age

The study of the hindsight effect in children is of particular difficulty, since the verbal methods used in experiments on adults are too complex for children's perception. Experimental procedures have been developed, including visual identification, to study this phenomenon in a child sample. The procedure begins with presenting the child with a blurry image, which becomes clearer over time. In some cases, the subject knows what the final object will be, but in others it does not. In cases where the child knows what is shown in the picture, he is asked to estimate how much time it will take other children of his age to guess what is shown in the picture. Due to the influence of the highsite error, the estimated time is usually much lower than the real guessing time, since the subjects in making the assessment rely on the information received from the researcher ^[16] .

The results of these studies show that children are subject to the highside effect in the same way as adults. The basis of the Hindsight error, both in adults and in children, is the same limitation of knowledge. This limitation is manifested in the person’s dependence on the actual knowledge available to him during an attempt to recall or reason about a more naive (relatively accessible information) position, regardless of whether it is a reflection on one’s own naive position or on the situation of another person. This key limitation that underlies the highsite error can explain other features and determining factors of children's cognition ^[17] .

The Role of Surprise

Experiencing surprise can partially explain those cases when a highsite error does not occur. Surprise affects the process of restoring in memory the expectations of certain events (results) that a person had before these events occurred. Initial surprise starts the process of comprehension. If the process of comprehension remains incomplete and sensory information has not been detected and encoded, this state is experienced as a surprise, which leads to a decrease in the effects of a highsite error. Thus, in a situation of insufficient understanding, the phenomenon of a “reverse highsite error” occurs ^[18] .

Memory Distortion

The highsite error is similar to other memory distortion phenomena, such as the disinformation effect and false memories of autobiographical memory ^[19] . All three memory distortion options are the result of a three-step process. ^{[19] The} features of each process for the three cases may vary, but they all end with a particular psychological manipulation or change of memory. The first stage is different for three phenomena, however, in all three cases, at the first stage there is a certain event: the event that occurred (disinformation effect); an event that was not (false autobiographical memories); a statement made by a person about an event that he remembers (a highsite error). The second stage is to increase the information that a person receives after an event has occurred. In the case of a highsite error, the new information is reliable and openly presented to the person, while the additional information in other cases of memory distortion is false and presented to the person in an ambiguous and possibly manipulative form. The third stage involves the restoration in memory (“remembering") of the original information. A person who is subject to a highsite error or the effect of misinformation should recall the initial information, while a person with false autobiographical memories will create memories of events that actually were not ^[19] .

Studies show that a person is inclined to make a highsite error even when he is aware of the existence of this effect and he has a desire to overcome it ^[20] . It is impossible to completely overcome the hindsight effect, however, there are ways to weaken it. One of them is the consideration of possible alternative explanations for the event and openness to various points of view ^[21] .

The only way to reduce the hindsight effect under experimental conditions is to make the subject think about alternative explanations that may be correct. As a result, the subject questions the correct (already confirmed) hypothesis and states that he could choose another alternative. Considering the fact that the researchers ’attempts to find ways to eliminate the highsite error failed, some authors believe that this effect is based on a combination of motivational and involuntary cognitive reconstruction processes ^[22] .