Hindsight bias: How knowledge and heuristics affect our reconstruction of the past

Once people know the outcome of an event, they tend to overestimate what could have been anticipated in foresight. Although typically considered to be a robust phenomenon, this hindsight bias is subject to moderating circumstances. In their meta-analysis, Christensen-Szalanski and Willham (1991) observed that the more experience people have with the task under consideration, the smaller is the resulting hindsight bias. This observation is one benchmark against which the explanatory power of process models of hindsight bias can be measured. Therefore, we used it to put the recently proposed RAFT model (Hoffrage, Hertwig, & Gigerenzer, 2000) to another test. Our findings were consistent with the "expertise effect." Specifically, we observed--using computer simulations of the RAFT model--that the more comprehensive people's knowledge is in foresight, the smaller is their hindsight bias. In addition, we made two counterintuitive observations: First, the relation between foresight knowledge and hindsight bias appears to be independent of how knowledge is processed. Second, even if foresight knowledge is false, it can reduce hindsight bias. We conclude with a discussion of the functional value of hindsight bias.     

Hindsight bias around the world

Hindsight bias refers to the tendency to overestimate in hindsight what one has known in foresight. Recently, two experiments extended the research to include samples from different cultures (Choi & Nisbett, 2000; Heine & Lehman, 1996). Asking their participants what they would have guessed before they knew the outcome ("hypothetical design"), Choi and Nisbett (2000) found that Koreans, in comparison to North Americans, exhibited more hindsight bias. Heine and Lehman (1996), however, reported that Japanese people in comparison to Canadians showed marginally less hindsight bias. In a second study, in which participants were asked to recall what they had estimated before they knew the outcome ("memory design"), the latter authors found no difference in hindsight bias between Japanese people and Canadians. We extended these studies with 225 Internet participants, in a hypothetical design, from four different continents (Asia, Australia, Europe, and North America). Hindsight bias was large and similar for all samples except for German and Dutch participants who showed no hindsight bias at all. While the latter effect may be based on peculiarities of the material and of the participants, the former underscores the worldwide stability of the phenomenon. In addition a follow-up surprise rating (paper and pencil) in China (35 participants) and Germany (20 participants) revealed that only less surprising items led to hindsight bias while more surprising ones did not. We suggest that the basic cognitive processes leading to hindsight bias are by-products of the evolutionary-evolved capacity of adaptive learning. On top of these basic processes, individual meta-cognitions (e.g., elicited by surprise) or motives (e.g., a self-serving motive) may further moderate the amount of bias, thus explaining the diverging results of Choi and Nisbett (2000), Heine and Lehman (1996), and our own study.     

Hindsight bias and causal reasoning: a minimalist approach

What factors contribute to hindsight bias, the phenomenon whereby the known outcome of an event appears obvious only after the fact? The Causal Model Theory (CMT) of hindsight bias (Nestler et al. in Soc Psychol 39:182–188, 2008a; in J Expl Psychol: Learn Mem Cog 34:1043–1054, 2008b; Pezzo in Mem 11:421–441, 2003; Wasserman et al. in Pers Soc Psychol Bull 17:30–35, 1991) posits that hindsight bias can occur when people have the opportunity to identify potential causal antecedents and evaluate whether they could have led to the outcome. Two experiments incorporating highly controlled minimalist scenarios supported the CMT. As predicted by the CMT, hindsight bias occurred when the causal factor explained the actual outcome better than the alternative outcome, and reverse hindsight bias occurred when the causal factor explained the alternative outcome better than the actual outcome. Moreover, we found new evidence that outcome knowledge alone was insufficient to elicit hindsight bias in the absence of a potential causal antecedent. Implications for future directions in hindsight bias research are discussed.     

Hindsight bias in insight and mathematical problem solving: Evidence of different reconstruction mechanisms for metacognitive versus situational judgments

This article presents two experiments that used insight and mathematical problems to investigate whether different factors would affect hindsight bias on metacognitive and situational judgments. In both studies, participants initially rated their likelihood of solving each problem within a certain amount of time (metacognitive judgments) and rated the importance of each component of the problem for finding the solution (situational judgments). Next, participants attempted to solve each problem. In Experiment 1, all participants were given solution feedback information, but in Experiment 2, participants were not given any solution feedback. After 1 week, participants were asked to recall their original judgments. Hindsight bias was assessed by comparing the initial with the final ratings. Insight problems and math problems showed different patterns of hindsight bias effects on the metacognitive and situational judgments. The results suggest that two competing models of hindsight effects are actually complementary explanations for judgment reconstruction on different types of judgment tasks.     

Hindsight bias redefined: It’s about time

Four experiments introduced a new conceptual and methodological approach to hindsight bias, traditionally defined as the tendency to exaggerate the a priori predictability of outcomes after they become known. By examining likelihood estimates rooted to specific time points during an unfolding event sequence (videos and short text stories), judged both in foresight and hindsight, we conceptualized hindsight bias as a contrast between two “inevitability curves,” which plotted likelihood against time. Taking timing into account, we defined three new indicators of accuracy: linear accuracy (how well hindsight judgments capture the linear trend of foresight judgments over time), rate accuracy (how well hindsight judgments reflect the slope of foresight judgments over time), and temporal accuracy (how well hindsight judgments specify the overall timing of the full envelope of foresight judgments). Results demonstrated that hindsight judgments showed linear and rate accuracy, but were biased only in terms of lack of temporal accuracy. The oft-used catchphrase “knew it all along effect” was found to be a misnomer, in that participants were well aware in hindsight that their earlier foresight judgments reflected uncertainty. The current research therefore points to one way in which retrospective judgments can be considered biased, yet simultaneously suggests that considerable accuracy exists when people render such judgments.     

Hindsight bias from 3 to 95 years of age

Upon learning the outcome to a problem, people tend to believe that they knew it all along (hindsight bias). Here, we report the first study to trace the development of hindsight bias across the life span. One hundred ninety-four participants aged 3 to 95 years completed 3 tasks designed to measure visual and verbal hindsight bias. All age groups demonstrated hindsight bias on all 3 tasks; however, preschoolers and older adults exhibited more bias than older children and younger adults. Multinomial processing tree analyses of these data revealed that preschoolers' enhanced hindsight bias resulted from them substituting the correct answer for their original answer in their recall (a qualitative error). Conversely, older adults' enhanced hindsight bias resulted from them forgetting their original answer and recalling an answer closer to, but not equal to, the correct answer (a quantitative error). We discuss these findings in relation to mechanisms of memory, perspective taking, theory of mind, and executive function.     

Hindsight bias demonstrated in the prediction of a sporting event

The authors examined hindsight bias in the context of a sporting event (Super Bowl XXXIII) with individuals who had previous knowledge of the hindsight bias-42 students who were currently enrolled in psychology classes that had previously covered hindsight bias. To control for extraneous variables, the authors analyzed how often the participants watched football, their gender, and whom they predicted to win; they found no differences between how often the participants watched football and whom they predicted to win. The hindsight bias existed for the prediction of the outcome for the Super Bowl, and the participants with previous knowledge were not immune to its effect. Results are discussed in relation to attribution theory, the reconstructionist view, and response bias view.     

Hindsight bias after receiving self-relevant health risk information: A motivational perspective

The phenomenon of hindsight bias was explored in the context of self-relevant health risk information. Participants in a community screening estimated their cholesterol level (foresight measure) before receiving positive or negative feedback based on their actual cholesterol level. Hindsight estimations were then assessed twice: once immediately after the feedback, and again several weeks later. While the unexpected positive feedback group showed no systematic recall bias, hindsight estimations of individuals receiving unexpectedly negative feedback showed a dynamic change over time. Immediately after the feedback, participants' recollection of their expected cholesterol level were shifted towards their actual cholesterol level (hindsight bias). In contrast, several weeks later, foresight estimations were recalled as less accurate than they had been (reversed hindsight bias). These data might reflect a change of the motivational focus from "hot affect" and fear control, which occur immediately after receiving negative feedback, to danger control, which occurs some time after the feedback, as proposed by the dual process model.     

Hindsight bias after receiving self-relevant health risk information: a motivational perspective

The phenomenon of hindsight bias was explored in the context of self-relevant health risk information. Participants in a community screening estimated their cholesterol level (foresight measure) before receiving positive or negative feedback based on their actual cholesterol level. Hindsight estimations were then assessed twice: once immediately after the feedback, and again several weeks later. While the unexpected positive feedback group showed no systematic recall bias, hindsight estimations of individuals receiving unexpectedly negative feedback showed a dynamic change over time. Immediately after the feedback, participants' recollection of their expected cholesterol level were shifted towards their actual cholesterol level (hindsight bias). In contrast, several weeks later, foresight estimations were recalled as less accurate than they had been (reversed hindsight bias). These data might reflect a change of the motivational focus from "hot affect" and fear control, which occur immediately after receiving negative feedback, to danger control, which occurs some time after the feedback, as proposed by the dual process model.     

Hindsight bias as a function of anchor distance and anchor plausibility

This study explored the influence of anchor distance on hindsight bias and how the subjective plausibility of different anchors moderates this relation. In addition to the standard memory design used in hindsight bias research, participants were asked to indicate the range of values for possible answers to difficult almanac questions. Varying anchor distance on the basis of each participant's individual range of possible answers showed (1) that anchor plausibility decreased with increasing anchor distance following a non-linear monotone function, (2) that size of hindsight bias initially increased with increasing anchor distance but, from a certain distance, started to decrease, and (3) that hindsight bias was found to be always higher for plausible than for implausible anchors.     

The role of autobiographical memory networks in the experience of negative emotions: how our remembered past elicits our current feelings

The present research examined the role of autobiographical memory networks on negative emotional experiences. Results from 2 studies found support for an active but also discriminant role of autobiographical memories and their related networked memories on negative emotions. In addition, in line with self-determination theory, thwarting of the psychological needs for competence, autonomy, and relatedness was found to be the critical component of autobiographical memory affecting negative emotional experiences. Study 1 revealed that need thwarting in a specific autobiographical memory network related to the theme of loss was positively associated with depressive negative emotions, but not with other negative emotions. Study 2 showed within a prospective design a differential predictive validity between 2 autobiographical memory networks (an anger-related vs. a guilt-related memory) on situational anger reactivity with respect to unfair treatment. All of these results held after controlling for neuroticism (Studies 1 and 2), self-control (Study 2), and for the valence (Study 1) and emotions (Study 2) found in the measured autobiographical memory network. These findings highlight the ongoing emotional significance of representations of need thwarting in autobiographical memory networks.     

Hindsight bias as a function of anchor distance and anchor plausibility

This study explored the influence of anchor distance on hindsight bias and how the subjective plausibility of different anchors moderates this relation. In addition to the standard memory design used in hindsight bias research, participants were asked to indicate the range of values for possible answers to difficult almanac questions. Varying anchor distance on the basis of each participant's individual range of possible answers showed (1) that anchor plausibility decreased with increasing anchor distance following a non-linear monotone function, (2) that size of hindsight bias initially increased with increasing anchor distance but, from a certain distance, started to decrease, and (3) that hindsight bias was found to be always higher for plausible than for implausible anchors.     

Negative affect and hindsight bias

The relationship between hindsight bias and individual differences in negative affect, or ‘dysphoria’, was investigated in a naturalistic study. In a first session, 76 undergraduates predicted their grades prior to a midterm exam. In a second session, after having received feedback from the exam, they attempted to recall their predictions and predicted their grades on a second exam. Dysphoria was associated with hindsight bias whether initial predictions were overly optimistic or pessimistic. In the former case, the standard hindsight bias was modal for the sample, whereas a ‘reverse hindsight’ bias was modal in the latter. This asymmetry suggests that hindsight bias is influenced by motivational or affective factors. The bias did not hinder the improvement of predictions between the first and second exams, and therefore could not have mediated a hypothesized adaptive learning impairment among dysphoric subjects.     

Discussion: On the future of our past

After integrating the main lines of argumentation developed in the special issue, this paper makes a balance of the question on the public role of the university. Furthermore, the paper focuses on the discussion underlying the articles collected in the issue. A distinction is made between “comprehending” the past and present of the university on the one hand and “justifying” the university and its public role on the other hand. In view of this distinction, it is argued that the question concerning the value of a (normative) idea of the university and its public role is related to the question of democracy. The question is, however, not only whether we live in a democracy, but whether we may still speak of a “commitment to democracy”. An inquiry into the meaningfulness of a (normative) “idea” of the university arises directly from this question. At this point, the viewpoint of Derrida is introduced to discuss the question of the democracy and university of the future.     

Mood and heuristics: the influence of happy and sad states on sensitivity and bias in stereotyping

The influence of mood states on the propensity to use heuristics as expressed in stereotypes was examined using signal detection statistics. Participants experienced happy, neutral, or sad moods and "remembered" whether names connoting race (African American, European American) belonged to social categories (criminal, politician, basketball player). Positive mood increased reliance on heuristics, indexed by higher false identification of members of stereotyped groups. Positive mood lowered sensitivity (d'), even among relative experts, and shifted bias (beta) or criterion to be more lenient for stereotypical names. In contrast, sad mood did not disrupt sensitivity and, in fact, revealed the use of a stricter criterion compared with baseline mood. Results support theories that characterize happy mood as a mental state that predisposes reliance on heuristics and sad mood as dampening such reliance.     

Hindsight bias doesn't always come easy: causal models, cognitive effort, and creeping determinism

Creeping determinism, a form of hindsight bias, refers to people's hindsight perceptions of events as being determined or inevitable. This article proposes, on the basis of a causal-model theory of creeping determinism, that the underlying processes are effortful, and hence creeping determinism should disappear when individuals lack the cognitive resources to make sense of an outcome. In Experiments 1 and 2, participants were asked to read a scenario while they were under either low or high processing load. Participants who had the cognitive resources to make sense of the outcome perceived it as more probable and necessary than did participants under high processing load or participants who did not receive outcome information. Experiment 3 was designed to separate 2 postulated subprocesses and showed that the attenuating effect of processing load on hindsight bias is not due to a disruption of the retrieval of potential causal antecedents but to a disruption of their evaluation. Together the 3 experiments show that the processes underlying creeping determinism are effortful, and they highlight the crucial role of causal reasoning in the perception of past events.     

Hindsight bias in economic expectations: I knew all along what I want to hear

Hindsight bias in economic expectations was investigated with particular focus on the moderating effects of attitudes. Stronger hindsight bias was expected for subjectively favorable economic developments. Six months before and after the introduction of the euro as the official book currency of the European Monetary Union, participants rated the probabilities of several economic developments. Results show that hindsight bias occurs with economic expectations and that it is moderated by attitudes. Euro supporters showed stronger hindsight bias for positive economic developments than for negative ones, whereas euro opponents showed the opposite pattern. The results support the notion that hindsight bias is a reconstruction bias in which self-serving tendencies can influence the reconstruction selectively for favorable and unfavorable outcomes.     

At the core: how we steer our course

"The force that through the green fuse drives the flower drives my green age..."--Dylan Thomas (1952). The beautiful evocative words of Dylan Thomas speak to the driving life force that pulses through all creation. So what, then, drives our organization, the American Group Psychotherapy Association (AGPA)? What underlying stream can we tap into to provide the energy, power, and vitality we need to go on being in a meaningful and productive fashion? And, what internal compass can we rely on to steer us in the right direction, to guide us in how we function as an organization? My purpose here is to consider how we formulate our objectives, determine our priorities, make decisions, and insure, as best we can, that we are pursuing that which is meaningful and essential, not veering drastically off course. To do this, I would like to focus briefly on the current organizational structure of AGPA, then examine the use of a strategic planning process, review our accomplishments, and lastly explore hopes and dreams for our future as the premier group psychotherapy organization in the United States.     

Why versus how often: Causal reasoning and the incidence of judgmental bias

When faced with the task of making a prediction or estimating a likelihood, it is argued that people often reason about the presence or absence and relative strength of possible causal mechanisms for the production of relevant outcomes. In so doing people rely on “causal cues” or properties of an inferential problem which indicate the nature of the particular causal processes which give rise to specific outcomes. It is hypothesized that causal cues, precisely because they focus attention and thought on specific causal mechanisms, can obscure the relevance of mathematical laws of probability and lead to statistically biased judgment. Two experiments were conducted. Their results support the hypothesis, showing that the incidence of the conjunction fallacy and the base rate fallacy depend on task-specific cues for causal reasoning.     

Cognitive bias modification: induced interpretive biases affect memory

Previous research has shown that it is possible to experimentally induce interpretive biases using ambiguous scenarios. This study extends past findings by examining the effects of cognitive bias modification for interpretation on subsequent scenario recall. Participants were trained to interpret emotionally ambiguous passages in either a positive or negative direction. Transfer of the training to novel scenarios was tested. After training, participants were also asked to recall details from these novel scenarios. The results indicate that the training was effective in inducing the intended group differences in interpretive bias. Importantly, participants exhibited memory biases that corresponded to their training condition. These results suggest that manipulating interpretive biases can result in corresponding changes in memory. Findings from this study highlight the importance of future research on the relation among cognitive biases and on the possibility of modifying cognitive biases in emotional disorders.