Realism of confidence in sensory discrimination: The underconfidence phenomenon

This paper documents a very pervasive underconfidence bias in the area of sensory discrimination. In order to account for this phenomenon, a subjective distance theory of confidence in sensory discrimination is proposed. This theory, based on the law of comparative judgment and the assumption of confidence as an increasing function of the perceived distance between stimuli, predicts underconfidence—that is, that people should perform better than they express in their confidence assessments. Due to the fixed sensitivity of the sensory system, this underconfidence bias is practically impossible to avoid. The results of Experiment 1 confirmed the prediction of underconfidence with the help of present-day calibration methods and indkated-a-good quantitative fit of the theory. The results of Experiment 2 showed that prolonged experience of outcome feedback (160 trials) had no effect on underconfidence. It is concluded that the subjective distance theory provides a better explanation of the underconfidence phenomenon than-do previous accounts in terms of subconscious processes.     

The role of individual differences in the accuracy of confidence judgments

Generally, self-assessment of accuracy in the cognitive domain produces overconfidence, whereas self-assessment of visual perceptual judgments results in underconfidence. Despite contrary empirical evidence, in models attempting to explain those phenomena, individual differences have often been disregarded. The authors report on 2 studies in which that shortcoming was addressed. In Experiment 1, participants (N= 520) completed a large number of cognitive-ability tests. Results indicated that individual differences provide a meaningful source of overconfidence and that a metacognitive trait might mediate that effect. In further analysis, there was only a relatively small correlation between test accuracy and confidence bias. In Experiment 2 (N = 107 participants), both perceptual and cognitive ability tests were included, along with measures of personality. Results again indicated the presence of a confidence factor that transcended the nature of the testing vehicle. Furthermore, a small relationship was found between that factor and some self-reported personality measures. Thus, personality traits and cognitive ability appeared to play only a small role in determining the accuracy of self-assessment. Collectively, the present results suggest that there are multiple causes of miscalibration, which current models of over- and underconfidence fail to encompass.     

Calibration of sensory and cognitive judgments: Two different accounts

The purpose of this study was to test predictions of two recent theories of realism of confidence. Ecological approaches to realism of confidence in one's general knowledge (Gigerenzer et al. , 1991; Juslin, in press a ; Björkman, in press) predict good calibration or, in the case of poor cognitive adjustment, overconfidence, within the cognitive domain. The subjective distance theory of confidence in sensory discriminations (Björkman et al. , 1992) predicts a pervasive underconfidence bias for sensory discriminations. Empirical data are reported showing that: (a) Calibration for sensory judgments is considerably poorer than calibration for well adapted cognitive judgements, a difference that can be entirely traced to underconfidence in the sensory domain. (b) While an initial overconfidence bias in the cognitive domain is removed by outcome feedback, the bias observed in sensory discriminations is unaffected even by a prolonged feedback session. It is suggested that the nature of confidence in sensory discriminations is different from the nature of confidence in cognitive judgments.     

Impact of esteem-related feedback on mood, self-efficacy, and attribution of success: Self-enhancement/self-protection

It is suggested that positive ego-enhancing and negative ego-diminishing feedback conditions concerning social perceptiveness can give rise to altered affective states concerning self-esteem that generalize to other areas of cognitive functioning in such a way that this feedback accompanies a systematic self-serving attributional bias of cognitive/intellectual performance. The results of 40 participants in an ego-enhancing and 45 in an ego-diminishing feedback condition indicate more attribution of problem-solving success by the first group to effort and to ease of the task; participants who received ego-diminishing feedback attributed their initial problem-solving success less to ease of the task. Other more direct evidence of self-protective strategies on the part of those in the ego-diminishing feedback condition includes lower success expecta y on retest and reduced volunteering for additional challenge. The results are interpreted as evidence of the impact of generalized esteem-related affect on future success expectancy, on risk-taking, and on the attribution of success when future replication is a boundary condition.     

The contributions of anchoring and past-test performance to the underconfidence-with-practice effect

In the underconfidence-with-practice effect, people's judgments of learning (JOLs) typically underestimate memory performance across multiple study-test phases. Whereas the past-test hypothesis suggests that this underconfidence stems from participants' reliance on earlier test performance to make subsequent JOLs (despite new learning), the anchoring hypothesis suggests that the underconfidence stems from participants' reliance on a fixed psychological anchor point low on the JOL scale to make their JOLs. To contrast the predictions of these hypotheses, we had college students study, make JOLs, and test over several dozen paired-associate items across two study-test phases. We parametrically manipulated the presence or absence of testing and judging within participants during Phase 1. Contrary to the past-test hypothesis, items tested during Phase 1 demonstrated less underconfidence during Phase 2 than did nontested items. Furthermore, participants did not increase JOLs from Phase 1 to Phase 2 for items that they had not recalled or for items that had not been tested at all, suggesting that the underconfidence stemmed largely from participants' overreliance on a psychological anchor point to make their JOLs. Past test performance, however, seems to be a major cue that participants use to adjust their JOLs away from the anchor, reducing underconfidence. This was most evident when we used a between-participants manipulation (Exp. 2) to cause our participants to anchor their JOLs either high or low on the JOL scale, producing differential underconfidence independent of any adjustment. Taken together, these results support the anchoring hypothesis over the past-test hypothesis for explaining underconfidence with practice.     

The Landslide Victory That Wasn't: The Bias Toward Consistency in Recall of Election Support1

Experiments that test for the judgmental bias that results from a preference for cognitive consistency often contain two threats to their internal validity. First, the subjects are asked to make judgments about themselves. Thus, the biases that result may be explained in terms of cognitive consistency or the motivation to see oneself in a positive light. Second, the decision subjects are asked to make is often difficult to verify objectively. The present research sought evidence in support of cognitive consistency using a methodology that avoided these two confounds. The context chosen was the tendency of perceivers to use the outcome of a group decision to make inferences about the magnitude of group members’ support for the outcome. The present experiment examined whether people in Richmond, Virginia, would use the outcome of a gubernatorial election to make decisions regarding the percentage of people in favor of the winning candidate. Although the winner won by less than one-half of one percent of the popular vote, we found that the subjects significantly overestimated the degree of voter support he received and underestimated the degree of support his opponent received. Moreover, this tendency was exacerbated over time.     

On the failure of cognitive ability to predict myside and one-sided thinking biases

Two critical thinking skills—the tendency to avoid myside bias and to avoid one-sided thinking—were examined in three different experiments involving over 1200 participants and across two different paradigms. Robust indications of myside bias were observed in all three experiments. Participants gave higher evaluations to arguments that supported their opinions than those that refuted their prior positions. Likewise, substantial one-side bias was observed—participants were more likely to prefer a one-sided to a balanced argument. There was substantial variation in both types of bias, but we failed to find that participants of higher cognitive ability displayed less myside bias or less one-side bias. Although cognitive ability failed to associate with the magnitude of the myside bias, the strength and content of the prior opinion did predict the degree of myside bias shown. Our results indicate that cognitive ability—as defined by traditional psychometric indicators—turns out to be surprisingly independent of two of the most important critical thinking tendencies discussed in the literature.     

The Role of Individual Differences in the Accuracy of Confidence Judgments

Generally, self-assessment of accuracy in the cognitive domain produces overconfidence, whereas self-assessment of visual perceptual judgments results in under-confidence. Despite contrary empirical evidence, in models attempting to explain those phenomena, individual differences have often been disregarded. The authors report on 2 studies in which that shortcoming was addressed. In Experiment 1, participants (N = 520) completed a large number of cognitive-ability tests. Results indicated that individual differences provide a meaningful source of overconfidence and that a metacognitive trait might mediate that effect. In further analysis, there was only a relatively small correlation between test accuracy and confidence bias. In Experiment 2 (N = 107 participants), both perceptual and cognitive ability tests were included, along with measures of personality. Results again indicated the presence of a confidence factor that transcended the nature of the testing vehicle. Furthermore, a small relationship was found between that factor and some self-reported personality measures. Thus, personality traits and cognitive ability appeared to play only a small role in determining the accuracy of self-assessment. Collectively, the present results suggest that there are multiple causes of miscalibration, which current models of over- and underconfidence fail to encompass.     

Moderators of Candidate Name‐Order Effects in Elections: An Experiment

Past studies of elections have shown that candidates whose names were listed at the beginning of a list on a ballot often received more votes by virtue of their position. This article tests speculations about the cognitive mechanisms that might be responsible for producing the effect. In an experiment embedded in a large national Internet survey, participants read about the issue positions of two hypothetical candidates and voted for one of them in a simulated election in which candidate name order was varied. The expected effect of position appeared and was strongest (1) when participants had less information about the candidates on which to base their choices, (2) when participants felt more ambivalent about their choices, (3) among participants with more limited cognitive skills, and (4) among participants who devoted less effort to the candidate evaluation process. The name‐order effect was greater among left‐handed people when the candidate names were arrayed horizontally, but there was no difference between left‐ and right‐handed people when the names were arrayed vertically. These results reinforce some broad theoretical accounts of the cognitive process that yield name‐order effects in elections.     

Effect of failure/success feedback and the moderating influence of personality on reward motivation

While motivation to pursue goals is often assumed to be a trait-like characteristic, it is influenced by a variety of situational factors. In particular, recent experiences of success or failure, as well as cognitive responses to these outcomes, may shape subsequent willingness to expend effort for future rewards. To date, however, these effects have not been explicitly tested. In the present study, 131 healthy individuals received either failure or success feedback on a cognitive task. They were then instructed to either ruminate or distract themselves from their emotions. Finally, they completed the Effort Expenditure for Rewards Task, a laboratory measure of reward motivation. Results indicate that participants who received failure feedback relied more strongly on the reward magnitude when choosing whether to exert greater effort to obtain larger rewards, though this effect only held under conditions of significant uncertainty about whether the effort would be rewarded. Further, participants with high levels of trait inhibition were less responsive to reward value and probability when choosing whether to expend greater effort, results that echo past studies of effort-based decision-making in psychological disorders.     

Gender Differences in the Self-Assessment of Accuracy on Cognitive Tasks

Generally, self-assessment of accuracy in the cognitive domain produces overconfidence, whereas self-assessment of the accuracy of visual perceptual judgments produces underconfidence. The possible effect of gender differences on these robust findings appears to be underinvestigated. In this paper, I report two studies that take a step toward redressing this shortcoming. In Study 1, a group of young adults (N =185) were presented with a test of General Knowledge and a visual perceptual task. The results indicated the typical over/underconfidence phenomena noted above, but, when analyzed by sex, indicated statistically significant differences; men were more confident than women on both tasks. In Study 2, participants (N =303) with a wider age range completed 4 tests of cognitive ability, which were drawn from the theory of fluid and crystallized intelligence. Results indicated that the tendency for men to express higher levels of confidence than women in the accuracy of their work appears to remain constant across the life-span. These findings are discussed in relation to self-concept and gender stereotyping.     

Gender differences in automatic in-group bias: why do women like women more than men like men?

Four experiments confirmed that women's automatic in-group bias is remarkably stronger than men's and investigated explanations for this sex difference, derived from potential sources of implicit attitudes (L. A. Rudman, 2004). In Experiment 1, only women (not men) showed cognitive balance among in-group bias, identity, and self-esteem (A. G. Greenwald et al., 2002), revealing that men lack a mechanism that bolsters automatic own group preference. Experiments 2 and 3 found pro-female bias to the extent that participants automatically favored their mothers over their fathers or associated male gender with violence, suggesting that maternal bonding and male intimidation influence gender attitudes. Experiment 4 showed that for sexually experienced men, the more positive their attitude was toward sex, the more they implicitly favored women. In concert, the findings help to explain sex differences in automatic in-group bias and underscore the uniqueness of gender for intergroup relations theorists.     

Subjective probability intervals: how to reduce overconfidence by interval evaluation

Format dependence implies that assessment of the same subjective probability distribution produces different conclusions about over- or underconfidence depending on the assessment format. In 2 experiments, the authors demonstrate that the overconfidence bias that occurs when participants produce intervals for an uncertain quantity is almost abolished when they evaluate the probability that the same intervals include the quantity. The authors successfully apply a method for adaptive adjustment of probability intervals as a debiasing tool and discuss a tentative explanation in terms of a naive sampling model. According to this view, people report their experiences accurately, but they are naive in that they treat both sample proportion and sample dispersion as unbiased estimators, yielding small bias in probability evaluation but strong bias in interval production.     

The Effect of Information Strength and Weight on Behavior in Financial Markets

Griffin and Tversky (1992) explain evidence of individual over- and underconfidence as resulting from attending too much to the strength (i.e., extremity) of information and not enough to the weight (i.e., statistical reliability) of information. We report two experiments that demonstrate how information strength and weight affect confidence, trading, prices, and wealth in laboratory markets. Our results indicate that information strength and weight affect individual over- and underconfidence and that market participants lack sufficient self-insight to avoid trading when they are biased. As a consequence, market prices are biased, and market participants with high-strength, low-weight information systematically transfer wealth to participants with low-strength, high-weight information.     

No Reduction in Hindsight Bias after Complete Information and Repeated Testing

In hindsight, people often claim to have known more than they actually did. This finding has been termed hindsight bias. We report two hindsight-bias experiments which yielded converging results with different methods. All participants first estimated numerical values as answers to difficult knowledge questions, were later given the solutions, and were finally asked to recall their first estimates. In an attempt to reduce the typically found bias, the participants of one group in Experiment 1 were informed in advance about the experimental design and the bias phenomenon, while the participants of another group were not. In Experiment 2 all participants received at the end individual feedback about their recall performance and were then subjected to the same experimental procedure again, twice in succession. The amount of hindsight bias, however, remained unaffected by either of these two manipulations. These findings support automatic processes as explanation for the observed bias and dismiss motivational accounts.     

Problem Solvers Adjust Cognitive Offloading Based on Performance Goals

When incorporating the environment into mental processing (cf., cognitive offloading), one creates novel cognitive strategies that have the potential to improve task performance. Improved performance can, for example, mean faster problem solving, more accurate solutions, or even higher grades at university.¹ Although cognitive offloading has frequently been associated with improved performance, it is yet unclear how flexible problem solvers are at matching their offloading habits with their current performance goals (can people improve goal‐related instead of generic performance, e.g., when being in a hurry and aiming for a “quick and dirty” solution?). Here, we asked participants to solve a cognitive task, provided them with different goals—maximizing speed (SPD) or accuracy (ACC), respectively—and measured how frequently (Experiment 1) and how proficiently (Experiment 2) they made use of a novel external resource to support their cognitive processing. Experiment 1 showed that offloading behavior varied with goals: Participants offloaded less in the SPD than in the ACC condition. Experiment 2 showed that this differential offloading behavior was associated with high goal‐related performance: fast answers in the SPD, accurate answers in the ACC condition. Simultaneously, goal‐unrelated performance was sacrificed: inaccurate answers in the SPD, slow answers in the ACC condition. The findings support the notion of humans as canny offloaders who are able to successfully incorporate their environment in pursuit of their current cognitive goals. Future efforts should be focused on the finding's generalizability, for example, to settings without feedback or with high mental workload.     

Calibration of sensory and cognitive judgments: A single model for both

In a recent issue of this journal, Winman and Juslin (34 , 135–148, 1993) present a model of the calibration of subjective probability judgments for sensory discrimination tasks. They claim that the model predicts a pervasive underconfidence bias observed in such tasks, and present evidence from a training experiment that they interpret as supporting the notion that different models are needed to describe judgment of confidence in sensory and in cognitive tasks. The model is actually part of the more comprehensive decision variable partition model of subjective probability calibration that was originally proposed in Ferrell and McGoey (Organizational Behavior and Human Performance, 26 , 32–53, 1980). The characteristics of the model are described and it is demonstrated that the model does not predict underconfidence, that it is fully compatible with the overconfidence frequently found in calibration studies with cognitive tasks, and that it well represents experimental results from such studies. It is concluded that only a single model is needed for both types of task.     

Compensation method affects risk-taking in the Balloon Analogue Risk Task

Different participant compensation methods may have discrepant effects on decision-making in behavioral measures of risk-taking. Participants in clinical samples tend to receive session-based payment (often in conjunction with decision-based payment), whereas participants in student samples generally receive decision-based payment or no payment at all. This study examined the effect of different methods of participant payment on a behavioral measure of individual differences in risk-taking. Participants completed the Balloon Analogue Risk Task (BART) as well as questionnaire measures of sensation-seeking and impulsivity. Participants who received session-based payment engaged in significantly greater risk-taking in the BART compared to those who were paid based on their decisions and those who were not paid at all (i.e., those who were only compensated with course credit). These effects were not influenced by age, gender, sensation-seeking or impulsivity. These findings provide evidence that different compensation methods significantly influence participants’ risk-taking propensity as measured by the BART.     

1/f noise and effort on implicit measures of bias

Phenomena that vary over time can often be represented as a complex waveform. Fourier analysis decomposes this complex wave into a set of sinusoidal component waves. In some phenomena, the amplitude of these waves varies in inverse relation to frequency. This pattern has been called 1/f noise and, unlike white noise, it reflects nonrandom variation. Latencies in simple computer tasks typically reveal 1/f noise, but the magnitude of the noise decreases as tasks become more challenging. The current work hypothesizes a correspondence between 1/f noise and effort, leading to the prediction that increasing effort will reduce 1/f noise. In 2 studies, the author examined the relationship between an individual's attempts to avoid bias (measured in Study 1, manipulated in Study 2) and 1/f noise in implicit measures of stereotyping and prejudice. In each study, participants who made an effort to modulate the use of racial information showed less 1/f noise than did participants who made less effort. The potential value of this analytic approach to social psychology is discussed.