No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
"I Knew It All Along" Eventually: The Development of Hindsight Bias in Reaction to the Clinton Impeachment Verdict
Abstract
Students estimated the chance that President Clinton would be convicted or acquitted in the U.S. Senate impeachment trial, at 4 points in time: (a) 22 days before the verdict (distal pretest), (b) 3 days before the verdict (proximal pretest), (c) 4 days after the verdict (immediate posttest), and (d) 11 days after the verdict (delayed posttest). Within-subjects analyses revealed that both foresight and hindsight changed over time. With respect to foresight, the perceived likelihood of conviction significantly decreased and the perceived likelihood of acquittal significantly increased between the 2 pretests. Content analyses of archived national news reports supported the notion that media predictions about the impending vote produced this temporal shift toward greater prospective accuracy. Confirming hypotheses, hindsight bias emerged for estimates of the prior probability of conviction, but not for estimates of the prior probability of acquittal. Within-subjects analyses revealed that 4 days after the verdict, participants accurately recalled that their prospective estimates of the likelihood of conviction had shifted toward greater accuracy over time; but 1 week later, they incorrectly believed that they had been more certain all along that Clinton would not be convicted. This article suggests that the potential losses associated with conviction were more salient than the potential gains associated with acquittal, and speculates that cognitive assimilation of the past did not occur until people had forgotten their earlier beliefs. These findings are important because they demonstrate that whether or not hindsight bias emerges depends not only on how long after the event people make retrospective judgments, but also how long before the event people focus these retrospective judgments. Supporting a self-serving interpretation, hindsight bias held for one's own estimates, but not for one's beliefs about the estimates of the average American or one's best friend. Consistent with Hawkins and Hastie's (1990) conclusions, a hybrid of cognitive and motivational accounts thus best explains the results.
... Studies were coded as real-world events if participants were asked to retrospectively judge the likelihood of an event that actually occurred in the recent past. An example of such a real-world event is the Senate impeachment trial of President Clinton (Bryant & Guilbault, 2002). Studies were coded as case histories if participants were given an example of the event and were asked to judge the likelihood of its occurrence. ...
... Studies were coded as to whether the outcome of the event was positive, negative, or neutral. Based on the notion that losses loom larger than gains (Kahneman & Tversky, 1979), it was predicted that hindsight bias would be stronger for negative events than for either positive or neutral events (Bryant & Guilbault, 2002). ...
... Although the time between pretest and posttest or between event and posttest was not found to be a significant moderator variable in this meta-analysis, recent investigations of real-world events indicate that this temporal dimension may influence the magnitude of hindsight bias. Specifically, studies of hindsight bias in reaction to the acquittal verdict in the O. J. Simpson murder trial (Bryant & Brockway, 1997) and in the U.S. Senate impeachment trial of President Clinton (Bryant & Guilbault, 2002) both found that hindsight bias increased after the outcome as a function of the passage of time. Clearly, more research is needed investigating how time affects the strength and trajectory of hindsight bias. ...
We conducted a meta-analysis of research on hindsight bias to gain an up-to-date summary of the overall strength of hindsight effects and to test hypotheses about potential moderators of hindsight distortion. A total of 95 studies (83 published and 12 unpublished) were included, and 252 independent effect sizes were coded for moderator variables in 3 broad categories involving characteristics of the study, of measurement, and of the experimental manipulation. When excluding missing effect sizes, the overall mean effect size was M d = .39 with a 95% confidence interval of .36 to .42. Five main findings emerged: (a) effect sizes calculated from objective probability estimates were larger than effect sizes calculated from subjective probability estimates; (b) effect sizes of studies that used almanac questions were larger than effect sizes of studies that used real-world events or case histories; (c) studies that included neutral outcomes resulted in larger effect sizes than studies that used positive or negative outcomes; (d) studies that included manipulations to increase hindsight bias resulted in significantly larger effect sizes than studies in which there were no manipulations to reduce or increase hindsight bias; and (e) studies that included manipulations to reduce hindsight bias did not produce lower effect sizes. These findings contribute to our understanding of hindsight bias by updating the state of knowledge, widening the range of known moderator variables, identifying factors that may activate different mediating processes, and highlighting critical gaps in the research literature.
... Indeed, studies that have focused on timinge.g. how long prior to a trial verdict it is when the outcome judgment is madehave shown that the further in the past an event had occurred could mean a relatively stronger bias (Bryant & Guilbault, 2002). One could predict, then, that the longer one has been an "expert," the more pointless it would be to try to counteract the hindsight bias, as the memory for not yet knowing may have completely decayed. ...
... Here, they judged only the 10 that were identified previously, perhaps making it more realistic to go back to their actual naïve state. As brought up in the introductionif more time has passed since learning, the naïve prior memories may be more difficult to remember, or even gone, resulting in a sure bias (Bryant & Guilbault, 2002). Thus, it may be that in Experiment 1, having to identify so many items made it easier to forget the past, while in Experiment 2, the fewer number of items not only made it okay to remember the naïve past, but conceivable. ...
Hindsight bias is a phenomenon that occurs when outcome knowledge interferes with the ability to accurately recall judgments made in a previous, naïve state. Also known as the “knew it all along” bias, we aimed to diminish the bias by having individuals take the perspective of a naïve other, as a way of encouraging acceptance that they had, in fact, not known it all along. Adult participants were given blurry-to-clear images incrementally until they were able to identify the object and were then re-presented with the same sequence of images and asked to make a judgment about when they had identified the item correctly the first time. They were also asked to judge when they thought a naïve peer (Experiments 1 and 2), or a naïve child (Experiment 2) could identify the objects. Results showed a robust hindsight bias in all perspectives, and sporadic success at eliminating the bias. When taking the perspective of a naïve peer, there were failures and successes; when taking the perspective of a naïve child, there was an ultra-debiasing, or a reverse hindsight bias. However, did the manipulation backfire? We conclude that while the manipulation of thinking like a naïve child may have eliminated the bias, participants seemed to use an “adults know best” rule rather than accepting past naivete for themselves.
... In hindsight, individuals judge events or outcomes as more foreseeable than they would have in foresight. This is a robust cognitive error that has been largely observed using almanac questions, real-world events, political election outcomes and natural disasters (e.g., Bryant & Guilbault, 2002;Roese & Vohs, 2012). The bias has also been found to increase with task difficulty (e.g., Chen et al., 2020;Fischhoff, 1977;Harley et al., 2004) and is difficult to eliminate (e.g., Bernstein et al., 2011). ...
Visual hindsight bias, also known as the "saw-it-all-along" effect, is the tendency to overestimate one's perceptual abilities with the aid of outcome knowledge. Recently, Giroux et al. (2022, Emotion, https://doi.org/10.1037/emo0001068 ) reported robust visual hindsight bias for emotional faces except for happy. We examined whether the difficulty of emotional processing could explain their finding. As in Giroux et al., participants saw a blurred image of an emotional face (happy, angry, or neutral) that progressed to clear and were instructed to stop the clearing process when they were able to identify the emotion (foresight trials). They then were shown the clearest image of each face and determined the emotion, followed by a memory task where they were asked to adjust the blur levels to indicate the point at which they had identified the emotion earlier (hindsight trials). Experiment 1 replicated Giroux et al.'s finding, showing that participants stopped the image at a higher degree of blur during the hindsight trials than they had during the foresight trials (i.e., a visual hindsight bias) for the angry and neutral faces but not happy faces. Experiment 2 manipulated the perceptual difficulty of angry and happy faces. While the easy faces replicated the results of Experiment 1, both angry and happy faces produced strong bias when made difficult. A multinomial processing tree model suggests that visual hindsight bias for emotional faces, while robust, is sensitive to perceptual processing difficulties across emotions.
... While the two procedures usually yield similar results (Guilbault et al., 2004), there is some evidence to suggest that the hindsight bias is stronger when it comes to elapsed times of days or weeks between the event and the recollection of the original judgments. This was found in studies investigating reactions to the verdict in the O. J. Simpson criminal trial (Bryant & Brockway, 1997), the Clinton impeachment verdict (Bryant & Guilbault, 2002), and the outcomes of political elections (Blank et al., 2003). Examining whether the size of the hindsight bias on JOLs is affected by the delay between the memory test and JOL recollection might be an interesting avenue for future research. ...
Hindsight bias describes people's tendency to overestimate how accurately they have predicted an event's outcome after obtaining knowledge about it. Outcome knowledge has been shown to influence various forms of judgments, but it is unclear whether outcome knowledge also produces a hindsight bias on Judgments of Learning (JOLs). Three experiments tested whether people overestimated the accuracy of their memory predictions after obtaining knowledge about their actual memory performance. In all experiments, participants studied 60 cue-target word pairs, made a JOL for each word pair, and tried to recall the targets in a cued-recall test. In Experiments 1a and 1b, people recollected their original JOLs after attempting to recall each target, that is, after they obtained outcome knowledge for all items. In Experiments 2 and 3, people recollected their original JOLs in a separate phase after attempting to recall half the targets so that they had outcome knowledge for some but not all items. In all experiments, recollected JOLs were closer to actual memory performance than original JOLs for items with outcome knowledge only. Thus, outcome knowledge produced a hindsight bias on JOLs. Our results demonstrate that people overestimate the accuracy of their memory predictions in hindsight.
... These biases have in common that they describe our inability to put ourselves in the shoes of less-informed others (or our past self). With hindsight bias, we think that we "knew it all along" and overestimate the prior predictability of historic events (Fischhoff, 1975;Fischhoff and Beyth, 1975), elections (Blank et al., 2003), verdicts (Bryant and Guilbault, 2002), experimental results (Slovic and Fischhoff, 1977), entrepreneurial success (Bukszar and Connolly, 1988), or how early a patient's condition could have been diagnosed (Arkes et al., 1981;Berlin, 2003). 1 A key mechanism of information projection is that such "creeping determinism" leads to systematically biased performance evaluations: evaluators who judge with hindsight systematically underestimate the competence of others. ...
This study investigates a source of comparative overconfidence, or overplacement, which occurs when people overestimate themselves relative to others. We present a simple application of information projection (Madarász, 2012) to show that hindsight bias can lead to overplacement and excessive willingness to compete. We run an experiment in which subjects choose between a competitive tournament and piece-rate compensation after observing some of their competitors' past performance. We exogenously manipulate whether subjects have ex post information about their competitors' past tasks (hindsight) or not (no hindsight). We find that hindsight bias generates overplacement and increases subjects' valuation of tournament participation by 19%. In line with theory, the additional tournament entry in the hindsight setting is driven by low-performing participants who should not have entered the tournament.
... Some researchers asked their participants to estimate the probabilities of unknown out comes (and later recall their prediction) for political events, for example, the visit of Pres ident Nixon to Beijing and Moscow in 1972 (Fischhoff & Beyth, 1975), the British-Gurka war in India in 1814 (Fischhoff, 1975), a near-riot in Atlanta, Georgia, in 1967(Fischhoff, 1975, the Clarence Thomas confirmation vote for the U.S. Supreme Court in 1991 (Diet rich & Olson, 1993), or the impeachment verdict on President Bill Clinton in 1999 (Bryant & Guilbault, 2002). All studies found that an outcome was given a higher probability after it was established than before it occurred, thus exhibiting hindsight bias. ...
... Construal-level theory predicts that the hindsight bias and, more generally, theory-driven confi dence in the inevitability of past events will increase over time. Supporting this prediction, Bryant and Guilbault ( 2002 ) found that hindsight about Ex-President Clinton's acquittal in the Monica Lewinski case increased from 4 days aft er the verdict to 11 days aft er the verdict. Obviously, more research is needed to examine this hypothesis in more detail and to specify the conditions under which it would occur. ...
This chapter argues that abstraction has evolved in the service of prediction. Construallevel theory (CLT) is used to examine how people bridge the past and the future by means of abstraction, or, in other words, how people make predictions by using abstract mental construals. In the framework of CLT, prediction, or bridging over time, is akin to mentally bridging over other psychological distances: understanding another person's perspective (bridging over social distance), constructing alternate worlds (bridging over hypotheticality), or understanding other places and taking a different spatial perspective (mentally bridging over spatial distance). The chapter first introduces CLT and then describes relevant research on prediction, also pointing, along the way, to open questions for further research. Finally, it examines the commonalities among predictions over different types of distance.
... We agree, but suggest that the advantage of outcome-related information derives in part from the accompanying metacognitive experience of ease. Similarly, the observation that hindsight bias increases with temporal distance between prediction and outcome (Blank, Fischer, & Erdfelder, 2003;Bryant & Guilbault, 2002) is presumably not only a function of what comes to mind, but of the relatively higher difficulty of recalling more distant pre-event information. ...
Theories of hindsight bias have emphasized the role of declarative information. However, accumulating evidence suggests that the production and reduction of hindsight bias involves an intricate interplay of declarative information and metacognitive experiences. These experiences are informative in their own right and may qualify or even reverse the implications of thought content. We summarize and integrate the available findings in terms of a new model of hindsight bias that includes the interplay between: (a) declarative information, (b) metacognitive experiences, (c) their informational value, and (d) the naive theories used to interpret experiences.
... Other studies have used real-world events or episodes with two or more possible outcomes (e.g., Bryant & Guilbault, 2002;Fischer & Budescu, 1995;Mazursky & Ofir, 1990). The participant's task is to estimate the probability for each of the possible outcomes of an event or episode. ...
Hindsight bias is a robust phenomenon; it has been found with different designs, materials, and measures. However, several methodological problems may hinder an adequate analysis and interpretation of results obtained in experimental studies of the effect. This article therefore systematizes and critically discusses relevant features of designs, materials, and types of feedback, as well as different operationalizations and indices of hindsight bias. In particular, the potential confound of recollection and reconstruction, which may lead to inadequate theoretical conclusions, is addressed.
... There was a relatively short interval between the foresight judgements and the hindsight recollection of these judgements. Studies have shown that memory distortion increases with longer intervals Bryant & Brockway, 1997;Bryant & Guilbault, 2002;Nestler et al., 2010), suggesting a longer interval would have increased the magnitude of hindsight bias. The present study also used a relatively small set of questions. ...
One component of hindsight bias is memory distortion: Individuals' recollections of their predictions are biased towards known outcomes. The present study examined the role of working memory in the memory distortion component of hindsight bias. Participants answered almanac-like questions, completed a measure of working memory capacity, were provided with the correct answers, and attempted to recollect their original judgements in two conditions: with and without a concurrent working memory load. Participants' recalled judgements were more biased by feedback when they recalled these judgements with a concurrent memory load and working memory capacity was negatively correlated with memory distortion. These findings are consistent with reconstruction accounts of the memory distortion component of hindsight bias and, more generally, with dual process theories of cognition. These results also relate the memory distortion component of hindsight bias with other cognitive errors, such as source monitoring errors, the belief bias in syllogistic reasoning and anchoring effects. Implications for the separate components view of hindsight bias are discussed.
... The retention intervals differed between these two studies: Hell et al. had a 1-week interval between providing and recalling original judgments, whereas the present study had an interval of 20 min. Although the most recent meta-analysis on hindsight bias revealed that the time between original judgments and their recall was not a moderator of hindsight bias across studies (Guilbault, Bryant, Brockway, & Posavac, 2004), individual studies have shown that the memory distortion component of hindsight bias increases as the interval between judgments and their recall increases (e.g., Bryant & Brockway, 1997;Bryant & Guilbault, 2002;Nestler et al., 2010). It is possible that foresight response time would be negatively correlated with hindsight bias with a longer interval between foresight judgment and hindsight recall of that judgment, although elaborate encoding has been shown to have a greater effect on memory immediately after encoding than after a 1-week interval (Thapar & McDermott, 2001). ...
One component of hindsight bias is memory distortion. This component is measured with a memory design, in which individuals answer questions, learn the correct answers, and recall their original answers. Hindsight bias occurs when participants' recollections are closer to the correct answers than their original judgments actually were. The present study used a memory design to examine the relationship between response time in recalling original answers and the magnitude of hindsight bias. In Experiment 1, participants' response times were negatively correlated with a hindsight bias index. In Experiment 2, half of the participants were instructed to recall their original judgments quickly and the other participants were instructed to take time to recall their judgments. The hindsight bias index was greater among rapidly responding participants than among delayed responding participants. These results, in conjunction with other findings, support a separate components view of hindsight bias. The memory distortion component of hindsight bias appears to occur quickly, and unbiased responding requires time for processing. This finding relates the memory distortion component of hindsight bias to other cognitive biases, such as the belief bias in syllogistic reasoning. The relationship of this hindsight bias component to dual-process models of cognition is discussed, and several avenues for additional research are suggested. (PsycINFO Database Record (c) 2012 APA, all rights reserved).
... Finally, one may examine longer-term judgments. Subjective ease may cue probability (Bryant & Guilbault, 2002), recency (Herzog, Hansen, & Wänke, 2007), or sense making (Pezzo, 2003). For example, by definition, if outcomes do not make sense it is hard to generate reasons for them (see Pezzo & Pezzo, 2007; Wilson & Gilbert, 2008). ...
Thinking about the past is critical to everyday experiences, but people are not unbiased when doing so. Feelings of subjective ease that accompany generating reasons for known or alternative outcomes influence hindsight bias. But people do not always make decisions immediately after thinking about issues. Three experiments demonstrated that generating versus reading earlier generated reasons has markedly different effects on judgments, with theoretical and practical implications. Inevitability judgments were consistent with feelings of ease when generating reasons, but with numbers of reasons (content) when later reading those reasons. Experiments 2 and 3 also found that feelings of ease can be reconstructed if people reconsider their feelings when initially generating reasons. Discussion centers on the operation of subjective ease and its role in understanding judgment and decision making.
... But simple hindsight bias (Fischhoff, 1975) explanations alone may not easily account for changes over time only after failures (Study 2) or greater distancing for important attributes and for self versus others (Wilson & Ross, 2001). Noteworthy is that recent research has found that hindsight biases also can change over time (Bryant & Guilbault, 2002). It is interesting to speculate about whether people's judgments of future events parallel that of the past; that is, future events that could threaten people's self-regard might seem farther away, and also perhaps less inevitable , than do equally distant events that could enhance self-regard. ...
Three studies tested the hypothesis that thoughts about alternatives become increasingly accessible over time, leading poor outcomes to feel subjectively farther away and less inevitable. This subjective temporal distance bias was obtained even though actual time since poor and good outcomes was identical. In Study 1, participants who recalled distant poor team outcomes thought of alternatives easily and outcomes felt farther away and less inevitable. Thoughts about outcomes were most easily accessible after good outcomes, which felt closer and more inevitable. In Study 2, with measures obtained immediately or at a later time on a negotiation task, changes over time occurred primarily for poor team outcomes. In Study 3, team performance on an investment task indicated it is whether alternatives are thought of easily, not thought content, that produces this effect. Discussion centers on temporal appraisals, other temporal biases, and teams.
We explored hindsight bias and retroactive pessimism related to the occupation of Sofia University in 2013–2014, using a memory design. We found partial evidence for hindsight bias for predictions of whether the occupation would achieve its goal (the government's resignation) but not for specific events related to the occupation, which were deemed unlikely from the start and later perceived not to have occurred. We did not find evidence for retroactive pessimism: Hindsight bias indices for the prediction of the government's resignation were not reliably associated with support for the occupation and disappointment with its outcome. We propose using a recall/reconstruct measure as a more rigorous test for retroactive pessimism, which has so far been demonstrated through reassessment.
When a romantic relationship ends, individuals often look back and wish they would have done things differently. What may seem clear in hindsight, however, is often unclear in foresight. We investigated the effects of outcome knowledge on individuals’ judgments of a dating couple. In Study 1 (N1 = 181 U.S. college students, N2 = 334 U.S. community adults), participants read about a couple with an uncertain relationship trajectory; then, experimental group participants received knowledge about the couple’s status six months down the road as broken up or still together, while control group participants received no outcome knowledge. Individuals who were told the dating couple broke up perceived that outcome as more likely and obvious than did individuals who were not given outcome knowledge or who were told the couple stayed together. In Study 2 (N1 = 262 U.S. college students, N2 = 333 U.S. community adults), participants in the experimental conditions received knowledge about the couple’s status six months later as broken up or engaged, while control group participants received no outcome knowledge. In both samples, outcome knowledge of a breakup had a negative effect on individuals’ judgments about the couple. Among community adults, but not among college students, outcome knowledge of an engagement positively affected judgments of the couple. We offer directions for future research and discuss the mechanisms by which hindsight bias might affect evaluations of our own and others’ relationships.
The occurrence of major life events is associated with changes in well-being and personality. To better understand these effects, it is important to consider how individuals perceive major life events. Although theories such as appraisal theory and affective adaptation theory suggest that event perceptions change over time and that these changes are relevant for personality and well-being, stability and change of perceived event characteristics have not been systematically examined. This article aims to fill this gap using data from a longitudinal study (N = 619 at T1). In this study, participants rated nine perceived characteristics of the same major life event up to five times within 1 year. We estimated rank-order and mean-level stabilities as well as intraclass correlations of these life event characteristics with continuous time models. Furthermore, we computed continuous time models for the stability of affective well-being and the Big Five personality traits to generate benchmarks for the interpretation of the stability of the life event characteristics. Rank-order stabilities for the life event characteristics were lower than for the Big Five, but higher than for affective well-being. Most of the variance in life event characteristics was explained by between-person differences. Furthermore, we found a significant mean-level increase for the life event characteristic change in world views and a significant decrease for extraordinariness. These mean-level changes are in line with the meaning-making literature and affective adaptation theory, whereas the rather high rank-order stability of the life event characteristics challenges the importance of reappraisal processes of major life events. (PsycInfo Database Record (c) 2021 APA, all rights reserved).
This chapter describes research on the hindsight bias, which is the tendency for people to reappraise events after the fact as if they had been known all along. It also addresses the question of how time influences hindsight bias. It begins by briefly summarizing the three dominant methodological paradigms researchers have used to study hindsight judgments, considering the influence of time in relation to each of these dominant experimental approaches and highlighting relations between foresight (a priori prediction) and hindsight (a posteriori reconstruction) within each research paradigm. Then, it considers prevailing theoretical models of the processes that underlie the development of hindsight bias over time. It concludes by comparing and contrasting the role time plays within each mediational model, constructing an integrative temporal model and suggesting potentially fruitful avenues for future research on the role of time in the development of hindsight bias. © 2006 by Lawrence J. Sanna and Edward C. Chang. All rights reserved.
Hindsight bias occurs when people feel that they "knew it all along," that is, when they believe that an event is more predictable after it becomes known than it was before it became known. Hindsight bias embodies any combination of three aspects: memory distortion, beliefs about events' objective likelihoods, or subjective beliefs about one's own prediction abilities. Hindsight bias stems from (a) cognitive inputs (people selectively recall information consistent with what they now know to be true and engage in sensemaking to impose meaning on their own knowledge), (b) metacognitive inputs (the ease with which a past outcome is understood may be misattributed to its assumed prior likelihood), and (c) motivational inputs (people have a need to see the world as orderly and predictable and to avoid being blamed for problems). Consequences of hindsight bias include myopic attention to a single causal understanding of the past (to the neglect of other reasonable explanations) as well as general overconfidence in the certainty of one's judgments. New technologies for visualizing and understanding data sets may have the unintended consequence of heightening hindsight bias, but an intervention that encourages people to consider alternative causal explanations for a given outcome can reduce hindsight bias.
© The Author(s) 2012.
Since Baruch Fischhoff's (1975) groundbreaking paper opened up a whole new research field, more than 150 journal articles and book chapters, two meta-analyses (Christensen-Szalanski & Willham, 1991; Guilbault, Bryant, Brockway & Posavac, 2004), and one special issue (Memory, 2003, edited by Ulrich Hoffrage & Rodiger Pohl) have addressed hindsight phenomena. The current editorial aims to provide a rough roadmap to the hindsight bias research landscape. It highlights some important Iandmarks and developments of the last 30 years and puts the 13 articles of the present special issue into a historical and systematic perspective.
The purpose of this paper is to determine how gender, party affiliation, political views, age, race, education, income, attendance at religious services, and other such variables affected public opinions of President Clinton shortly after the release of the Starr report (and before the impeachment vote in the House of Representatives). Using data from a CBS NEWS / NEW YORK TIMES poll taken from September 12 to September 15, 1998, this study found the following: 1) Age had an effect on opinions concerning the Clinton scandal with younger individuals more likely to want the impeachment process to begin. This information provides some support for the Life-experience hypothesis. 2) Women were more likely to want Clinton to resign and less likely to want impeachment, which is perhaps consistent with the strong support Clinton had from women voters in the 1996 election. 3) Those with more education and income were more likely to want Clinton to resign and less likely to want impeachment, which is consistent with the resource hypothesis. 4) The combination of all attributes and individual characteristics determine how final opinions are established concerning resignation, impeachment, and dropping the matter.
This chapter contains section titled:
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.
The answer is three: questioning a conceptual default assumption in hindsight bias research, we argue that the hindsight bias is not a unitary phenomenon but consists of three separable and partially independent subphenomena or components, namely, memory distortions, impressions of foreseeability and impressions of necessity. Following a detailed conceptual analysis including a systematic survey of hindsight characterizations in the published literature, we investigated these hindsight components in the context of political elections. We present evidence from three empirical studies that impressions of foreseeability and memory distortions (1) show hindsight effects that typically differ in magnitude and sometimes even in direction, (2) are essentially uncorrelated, and (3) are differentially influenced by extraneous variables. A fourth study found similar dissociations between memory distortions and impressions of necessity. All four studies thus provide support for a separate components view of the hindsight bias. An important consequence of such a view is that apparent contradictions in research findings as well as in theoretical explanations (e.g., cognitive vs. social-motivational) might be alleviated by taking differences between components into account. We also suggest conditions under which the components diverge or converge.
It was hypothesized that hindsight effects and biased causal attributions both result from the causal explanations that perceivers construct when linking antecedents to outcomes at encoding. Specifically, the likelihood of a particular outcome and its perceived causal origins should depend on the number of causal antecedents available, the complexity of subjects' outcome explanations, and the number of outcome alternatives considered. After subjects were exposed to variations in event-specific information, they either learned the outcome or not and then explained one or more outcome alternatives. As predicted, estimates of outcome likelihood were inflated as the complexity of subjects' causal explanations increased. However, inducing subjects to explain alternative outcomes-particularly several alternative outcomes-reduced their tendency to exaggerate the likelihood of the known outcomes. Similarly, causal attributions varied as a function of the outcome alternatives explained and the number of causal antecedents presented. Implications for legal strategies are discussed.
Investigated how often people are wrong when they are certain that they know the answer to a question. Five studies with a total of 528 paid volunteers suggest that the answer is "too often." For a variety of general-knowledge questions, Ss first chose the most likely answer and then indicated their degree of certainty that the answer they had selected was, in fact, correct. Across several different question and response formats, Ss were consistently overconfident. They had sufficient faith in their confidence judgments to be willing to stake money on their validity. The psychological bases for unwarranted certainty are discussed in terms of the inferential processes whereby knowledge is constructed from perceptions and memories. (15 ref) (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Judges who had estimated the likelihood of various possible outcomes of President Nixon's trips to Peking and Moscow were unexpectedly asked to remember, or reconstruct in the event that they had forgotten, their own predictions some time after the visits were completed. In addition, they indicated whether or not they thought that each event had in fact occurred. Remembered—reconstructed probabilities were generally higher than the originally assigned probabilities for events believed to have occurred and lower for those which had not (although the latter effect was less pronounced). In their original predictions, subjects overestimated low probabilities and underestimated high probabilities, although they were generally quite accurate. Judging by their reconstructed—remembered probabilities, however, subjects seldom perceived having been very surprised by what had or had not happened. These results are discussed in terms of cognitive “anchoring” and possible detrimental effects of outcome feedback.
Other studies have reported that when subjects are presented with outcome feedback they are unable to remember their original knowledge state (the “knew-it-all-along effect”). In the present studies, feedback was followed by manipulations which were intended to invalidate it. In the first experiment, we failed to discredit the feedback and so report a knew-it-all-along effect under circumstances different from those reported elsewhere. In the second experiment, the discrediting instructions were successful and the effect was disrupted. Contrary to previous interpretations, the latter results indicate that feedback information is not automatically assimilated and that people can access their prior knowledge state, if the circumstances require.
Tested the hypothesis that an individual will feel control over an outcome if he causes the outcome and if he knows before causing it what he hopes to obtain. 65 male undergraduates were shown 2 consumer items and told that they would get to win 1 by a chance drawing. 2 marbles of different colors were placed in a can and mixed up. One-third of the Ss were told that the E would pick a marble to determine their prize and were told beforehand which marble stood for which prize. Another third were told to select a marble to determine their prize and were told beforehand which marble stood for which prize. The remaining Ss were told to select a marble to determine their prize but were not told until after they had picked their marble which marble stood for which prize. Ss then received a marble which led them to win either the item they preferred or the item they did not prefer. Results strongly support the hypothesis: Ss who caused their own outcome and knew beforehand what they hoped to obtain perceived themselves to have more control over the outcome, more choice about which outcome they received, and more responsibility for their outcome than Ss in the remaining conditions. These results were replicated in a 2nd experiment with the same Ss. The relationship between these studies and previous experiments on control is explored, and some possibilities for future research on control are considered. (23 ref) (PsycINFO Database Record (c) 2006 APA, all rights reserved).
If subjects are asked to recollect a former response after having been informed about the correct response, their recollection
tends to approach the correct response. This effect has been termedhindsight bias. We studied hindsight bias in an experiment requiring numerical responses to almanac-type questions for physical quantities.
We varied (1) the time at which the correct information was provided, (2) the encoding of the original responses by asking/not
asking subjects to give a reason for the respective response, and (3) the motivation to recall correctly. We found that hindsight
is less biased if reasons are given and if the correct information is provided at an earlier time. Motivation had only interactive
effects: (1) With high motivation to recall correctly, the time the correct information was provided had no influence. (2)
With reasons given, the variation of motivation showed no effect. These results rule out purely motivational and purely automatic
explanations.
CONSIDERS EXPERIMENTAL RESEARCH THAT HAS USED PROBABILITY THEORY AND STATISTICS AS A FRAMEWORK WITHIN WHICH TO STUDY HUMAN STATISTICAL INFERENCE. EXPERIMENTS HAVE INVESTIGATED ESTIMATES OF PROPORTIONS, MEANS, VARIANCES, AND CORRELATIONS, BOTH OF SAMPLES AND OF POPULATIONS. IN SOME EXPERIMENTS, PARAMETERS OF POPULATIONS WERE STATIONARY; IN OTHERS, THE PARAMETERS CHANGED OVER TIME. THE EXPERIMENTS ALSO INVESTIGATED THE DETERMINATION OF SAMPLE SIZE AND TRIAL-BY-TRIAL PREDICTIONS OF EVENTS TO BE SAMPLED FROM A POPULATION. IN GENERAL, THE RESULTS INDICATE THAT PROBABILITY THEORY AND STATISTICS CAN BE USED AS THE BASIS FOR PSYCHOLOGICAL MODELS THAT INTEGRATE AND ACCOUNT FOR HUMAN PERFORMANCE IN A WIDE RANGE OF INFERENTIAL TASKS. (115 REF.) (PsycINFO Database Record (c) 2006 APA, all rights reserved).
Individuals who have not been victimized by negative life events, such as serious illness, accidents, or crime, tend to perceive themselves as “uniquely invulnerable,” as less vulnerable to victimization than others. The actual experience of victimization, however, appears to shatter this illusion of invulnerability, creating in victims a new and unfamiliar sense of vulnerability often accompanied by psychological distress. This article reviews literature documenting nonvictims' perceptions of unique invulnerability and victims' heightened perceptions of vulnerability, and addresses the potentially adaptive versus maladaptive consequences of these perceptions. It is argued that victims who have the most difficulty coping with their misfortune may be precisely those individuals who initially felt least vulnerable prior to being victimized. Therefore, how victims cope may depend in part on their prior beliefs about risk. In addition, a distinction is made between victims who feel “uniquely vulnerable” (more vulnerable than others) and those who feel “universally vulnerable” (equally vulnerable as others) to future misfortune. It is proposed that perceptions of universal vulnerability may be a more adaptive reaction to victimization than are perceptions of unique vulnerability.
The tendency for people retrospectively to overestimate the degree to which they expected certain events to occur was examined within the context of the 1980 presidential election. Previous research has concluded that distorted hindsight occurs due to people's inability to reconstruct prior probabilities for an event after it has occurred, but the possible mediation of motivational factors, specifically self-esteem and self-presentation, has not been adequately examined. Subjects were asked either before or after the 1980 presidential election, and under public or private response conditions, to predict the outcome of the election (preelection) or to indicate what they would have predicted the outcome to be had they been asked before the election (postelection). In addition, subjects were classified as being either high or low in ego involvement regarding knowledge of politics. Results showed clear evidence of hindsight distortion: Subjects asked after the election said they woud have predicted an outcome closer to the results of the election than those asked before, but there was no evidence of mediation by self-esteem or self presentation concerns.
Students estimated the chance that O.J. Simpson would be convicted or acquitted of murder and the likelihood of large-scale violence following conviction, at 3 points in time: 2 hr before the announcement of the not guilty verdict, 2 days after the verdict, and 1 week after the verdict. Within-subjects analyses revealed that estimates of the probability of conviction made 2 hr before the not guilty verdict were higher than estimates of prior probability made 2 days after the verdict. Supporting hypotheses, the tendency to recall a higher chance of conviction after the verdict was strongest among those who initially perceived acquittal to be least likely. Contrary to predictions, however, the perceived probability of acquittal did not significantly increase until 1 week after the verdict. Estimates of the prior likelihood of violence following a guilty verdict also increased immediately after the verdict and remained high 1 week later. Results are discussed in terms of the influence of mass media and cognitive processes on the development of hindsight bias over time.
in short, multivariate analysis of variance (MANOVA) is used to assess the statistical significance of the effect of 1 or more independent variables on a set of 2 or more dependent variables / begin . . . with an example of a research situation in which a MANOVA is used, followed by a discussion of some basic statistical concepts and the general purpose of a MANOVA / the assumptions underlying a MANOVA, as well as the consequences of violating those assumptions, are discussed / [explain] how a MANOVA is performed and how it relates to a traditional analysis of variance (ANOVA) / the ways in which MANOVA results are presented in journal articles and various analytical methods that often follow multivariate significance are reviewed / the latter sections . . . are devoted to multivariate analysis of covariance (MANCOVA), repeated measures MANOVA, and power analysis (PsycINFO Database Record (c) 2012 APA, all rights reserved)
self-identification approach to the analysis of stress
approach is transactional, in that it regards the phenomenon in terms of a dynamic interplay, or transaction, between people and the environmental forces that might change them (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Asked 144 male Ss to work in 4-person problem-solving groups. After providing recommendations for group solutions to a series of problems, Ss learned that they had been consistently in either the group's majority or minority—where majority decisions were binding on the entire group—and that either few (low dissent) or many (high dissent) group members had disagreed with the majority. Ss were then informed that their group had done extremely well, average, or very poorly on the problems. Majority Ss verdically rated their personal performance and personal responsibility for the group's performance more highly than did minority Ss except when the group failed. Under failure conditions, majority Ss defensively took as little responsibility as minority Ss and rated their personal performances equal to those of minority Ss. Ss in failing groups also defensively assigned less responsibility for the group's poor performance to themselves than they assigned to the poorest group member. The results support a self-serving bias explanation for egocentrism and fail to support predictions derived from a logical information-processing model. (39 ref) (PsycINFO Database Record (c) 2012 APA, all rights reserved)
When individuals learn the outcome of an event or the correct answer to a question, they overestimate its prior predictability: that is, they tend to believe they “knew it all along.” Cognitive and motivational interpretations of hindsight bias are briefly reviewed and a study designed to test the motivational interpretation is reported. Specifically, it was hypothesized that individual differences in the strengths of two motives, a need for predictability and a self-presentation motive, should be positively related to individual differences in the magnitude of hindsight bias. Sixty-eight subjects completed a Dogmatism Scale and an Intolerance for Ambiguity Scale (the predictability motive) and the Marlowe-Crowne Social Desirability Scale (the self-presentation motive) before participating in a standard hindsight-bias paradigm. Measures of both motives, as well as a self-reported ego-involvement measure, were positively associated with the amount of hindsight bias exhibited. Implications of this result for interpretations of hindsight and other conceptually related phenomena are discussed.
This paper considers the choice between an all-or-nothing (AON) rule and a proportionate-damages (PD) rule in civil litigation. Under AON, a prevailing plaintiff receives a judgment equal to his entire damages. Under PD, damages are reduced to reflect uncertainty. For example, if the trier of fact finds that there is a 75 percent chance that the defendant is liable, the judgment would equal 75 percent of the plaintiff's damages. Using a moral hazard model that takes into account defendants' decisions to comply with legal rules, evidentiary uncertainty, and settlement, we show that AON usually maximizes the rate of compliance, although it may result in a higher level of litigation. This, in turn, provides an efficiency rationale for the ubiquity of AON in the legal system. (c) 2009 by The University of Chicago. All rights reserved..
Analysis of decision making under risk has been dominated by expected utility theory, which generally accounts for people's actions. Presents a critique of expected utility theory as a descriptive model of decision making under risk, and argues that common forms of utility theory are not adequate, and proposes an alternative theory of choice under risk called prospect theory. In expected utility theory, utilities of outcomes are weighted by their probabilities. Considers results of responses to various hypothetical decision situations under risk and shows results that violate the tenets of expected utility theory. People overweight outcomes considered certain, relative to outcomes that are merely probable, a situation called the "certainty effect." This effect contributes to risk aversion in choices involving sure gains, and to risk seeking in choices involving sure losses. In choices where gains are replaced by losses, the pattern is called the "reflection effect." People discard components shared by all prospects under consideration, a tendency called the "isolation effect." Also shows that in choice situations, preferences may be altered by different representations of probabilities. Develops an alternative theory of individual decision making under risk, called prospect theory, developed for simple prospects with monetary outcomes and stated probabilities, in which value is given to gains and losses (i.e., changes in wealth or welfare) rather than to final assets, and probabilities are replaced by decision weights. The theory has two phases. The editing phase organizes and reformulates the options to simplify later evaluation and choice. The edited prospects are evaluated and the highest value prospect chosen. Discusses and models this theory, and offers directions for extending prospect theory are offered. (TNM)
A not very good film
- M Greenfield
Democrats urge censure measure but G.O.P. balks: Maneuvering heats up. The New York Times
- E Schmitt
Senators seek Lewinsky tape: Prosecutors question Jordan
- L Douglass
Rebuked but not removed? Senate Republicans seek alternative punishment
- G Tzemach
Rational choice and the framing of decisions. Decision making: Descriptive, normative, and prescriptive interactions
- A Tversky
- D Kahneman
Identity and self-identification
- B R Schlenker
Reading and understanding multivariate statistics
- K P Weinfurt