The dilution effect: judgmental bias,conversational convention,or a bit of both?

This study explored competing normative interpretations of the dilution effect: the tendency for people to underutilize diagnostic evidence in prediction tasks when that evidence is accompanied by irrelevant information. From the normative vantage point of the intuitive statistician, the dilution effect is a judgmental bias that arises from the representativeness heuristic (similarity-matching of causes and effects). From the normative prospective of the intuitive politician, however, the dilution effect is a rational response to evidence presented in a setting in which Gricean norms of conversation are assumed to hold. The current experiment factorially manipulated conversational norms, the degree to which diagnostic evidence was diluted by irrelevant evidence, and the accountability of subjects for their judgments. Accountable subjects demonstrated a robust dilution effect when conversational norms were explicitly primed as well as in the no-priming control condition, but no dilution when conversational norms were explicitly deactivated. Non-accountable subjects demonstrated the dilution effect across norm activation conditions, with the strongest effect under the activation of conversational norms. Although the results generally support the conversational-norm interpretation of dilution, the significant dilution effect among non-accountable subjects in the norm-deactivated condition is more consistent with the judgmental-bias interpretation.     

Does Feedback Improve the Accuracy of Recurrent Judgmental Forecasts?

Feedback has been shown to be a useful tool for improving decision making (Balzeret al.,1992) and might also be a useful tool for improving the accuracy of recurrent judgmental forecasts. The objective of this study was to examine the impact of feedback on accuracy when forecasting time series with structural instabilities. We found that task information feedback (prompting on the underlying structure of the time series) gave significantly better forecasting performance than performance outcome feedback (prompting with graphical indicators of forecasting accuracy or prompting words expressing levels of forecasting accuracy). We also found that adding cognitive information feedback (prompting on desirable forecasting behaviors) to task information feedback did not significantly improve forecasting performance. Task information and task information feedback with added cognitive information feedback, but not performance outcome feedback, were superior to the baseline of providing simple outcome feedback (following each forecast with the actual value of the time series).     

Interpersonal orientation and the accuracy of personality judgments

Are those who are more invested in developing and maintaining interpersonal relationships able to provide more accurate judgments of others' personality characteristics? Previous research has produced mixed findings. In the present study, a conceptual framework was presented and methods were used that overcome many of the problems encountered in past research on judgmental accuracy. On four occasions, 102 judges watched a 12‐min videotaped dyadic interaction and described the personality of a designated target person. Judges' personality characteristics were described by self, parents, and friends. Results revealed that psychological communion was positively associated with judges' accuracy in rating targets' personality characteristics. In addition, whereas women were more communal and provided more accurate judgments than men, the relationship between communion and accuracy held after controlling for the effect of gender. Finally, preliminary findings suggested that interpersonally oriented individuals may sometimes draw on information about themselves and about stereotypical others to facilitate accurate judgments of others.     

Improving judgmental forecasts with judgmental bootstrapping and task feedback support

This study examines the utility of two widely advocated methods for supporting judgmental forecasts—providing task feedback and providing judgmental bootstrapping support. In a simulated laboratory based experiment that focused on producing composite sales forecasts from three individual components, we compared the effectiveness of these two methods in improving final judgmental forecasts. In the presence of cognitive feedback task, feedback led to better forecasts than providing judgmental bootstrap forecasts. Simply providing bootstrap forecasts was of no additional benefit over a control condition. This was true in terms of the Brunswik Lens model measures of achievement, knowledge, and consistency, and in terms of forecast accuracy. This occurred both in stable environments and when special events (unusual one‐time events requiring adjustments to the forecasts) arose. Copyright © 2005 John Wiley & Sons, Ltd.     

Pure feedback effects in absolute identification

To reveal the pure effects of trial-by-trial feedback on judgmental accuracy and sequential dependencies independent of global anchoring effects and other influences, we presented subjects with sequences consisting alternately (within an experimental session) of short runs of trials with feedback (feedback sequences) and without feedback (no-feedback sequences). In Experiments 1 and 2 (absolute identification of sound intensity and sound frequency, respectively), judgmental accuracy was the same in the feedback and the no-feedback sequences, contrary to previous results. Also, in the feedback sequences, the dependency of the current response on the immediately preceding stimulus was larger than that in the no-feedback sequences, while the dependency on the previous response was larger in the no-feedback sequences. In Experiment 3 (absolute identification of sound frequency), we attempted to separate the effects of the number of response categories on sequential dependencies from the effects of the number of stimuli. The results showed that the number of response categories had a larger effect than the number of stimuli on most aspects of performance, but that both affected sequential dependencies. These results are generally consistent with a theory of absolute identification in which feedback affects judgmental accuracy by improving long-term memory for judgmental anchors, while feedback affects sequential dependencies by altering response biases.     

Modularity in everyday life judgments of aggression and violent behavior

This review focuses on one aspect of moral judgment of aggression and violent behavior in the context of the psychodynamics of everyday life: Judgmental modularity. The central hypothesis asserts that, from the victim's perspective, the severity of judgments or the relative weight assigned to physical damage, when information on intent and damage is available, will be maximized, whereas inverse trends will typify the judgments of the same person from the assailant's perspective. This view resembles the spirit of the functional approach to moral judgment of violent behavior. In this light, related studies that were conducted within the framework of functional measurement are reviewed. Judgmental modularity was documented in the majority of the findings. However, in two studies, the same participants exhibited judgmental consistency in the first phase and judgmental modularity in a second phase, which manipulated other types of judgmental perspectives. Implications for the issue of judgmental modularity, for the issue of modularity in violent behavior and for a proposal to establish a functional definition of aggression are discussed.     

Accountability: a social magnifier of the dilution effect

This research demonstrated that accountability can not only reduce judgmental bias, but also exacerbate it--in this case, the dilution effect. Ss made predictions from either diagnostic information alone or diagnostic information plus mixtures of additional data (nondiagnostic information, additional diagnostic data pointing to either the same conclusion or the opposite conclusion). Relative to unaccountable Ss, accountable Ss (a) diluted their predictions in response to nondiagnostic information and (b) were more responsive to additional diagnostic information. The accountability manipulation motivated subjects to use a wide range of information in making judgments, but did not make them more discriminating judges of the usefulness of that information.     

Feedback as a source of information and as a source of noise in absolute judgments of loudness.

Two experiments investigated the effects of feedback on absolute judgements of loudness. In Experiment 1, subjects received aceurate, unreliable, or no feedback. While feedback improved the information transmitted in judgments, it gave lower d' values than no feedback. These results were not compatible with a signal detection model with a noisy sensory stage and a decision stage with a fixed criterion, but suggested that criteria move in response to feedback and thus contribute judgmental noise to perceptual processes. Further confirmation for a variable criterion was obtained in Experiment 2, where reliability of feedback was held constant, but feedback was biased to favor some response alternatives more than others. Biased feedback shifted the positions of criteria, but also increased the inertia of some criteria in responding to feedback which caused changes in d'.     

Relatively certain! Comparative thinking reduces uncertainty

Comparison is one of the most ubiquitous and versatile mechanisms in human information processing. Previous research demonstrates that one consequence of comparative thinking is increased judgmental efficiency: comparison allows for quicker judgments without a loss in accuracy. We hypothesised that a second potential consequence of comparative thinking is reduced judgmental uncertainty. We examined this possibility in three experiments using three different domains of judgment and three different measures of uncertainty. Results consistently demonstrate that procedurally priming participants to rely more heavily on comparative thinking during judgment induces them to feel more certain about their judgment.     

The effects of source and type of feedback on child witnesses' metamemory accuracy

This study investigated the effect of feedback on the accuracy (realism) of 12‐year‐old children's metacognitive judgments of their answers to questions about a film clip. Two types of judgments were investigated: confidence judgments (on each question) and frequency judgments (i.e. estimates of overall accuracy). The source of feedback, whether it was presented as provided by a teacher or a peer child, did not influence metacognitive accuracy. Four types of feedback were given depending on whether the participant's answer was correct and depending on whether the feedback confirmed or disconfirmed the child's answer. The children showed large overconfidence when they received confirmatory feedback but much less so when they received disconfirmatory feedback. The children gave frequency judgments implying that they had more correct answers than they actually had. No main gender differences were found for any of the measures. The results indicate a high degree of malleability in children's metacognitive judgments. Copyright © 2004 John Wiley & Sons, Ltd.     

Nested models of relative judgment: Applications to a similarity averaging model

A theoretical analysis of the Eisler and Ekman (1959) model of similarity judgments for unidimensional continua is presented, based on a general model of relative judgment. This general model assumes that judgments are mediated by perceived relations of pairs of stimuli, that there exists a transformation of the judgmental response that is a function of the sensory ratio of the two stimuli, and that response bias operates in a multiplicative manner. Three structural conditions are presented, each imposing constraints on the structure of observed judgments. The structural conditions define threenested models of relative judgment, with the second a weakened version of the first, and the third a weakened version of the second. The special virtue of the general model is that it is applicable to a variety of judgmental tasks (e.g., ratio estimation, similarity, pair comparison), the key being derivation of theresponse transformation conforming to the structural conditions. The structural conditions thus constitute necessary conditions for several different judgmental models. The theory was first applied with success to ratio estimation judgments (Fagot, 1978), and this paper applies the general model to the Eisler and Ekman similarity “averaging” model. Empirical tests were carried out on published data for pitch, darkness, visual area, and heaviness judgments. Although the strong form of the model presented by Eisler and Ekman was rejected, weakened versions were generally supported by the data. These results were similar to those obtained for ratio estimation (Fagot, 1978), and are interpreted to be very promising for the general model of relative judgment.     

THE TRUTH-DECEPTION ATTRIBUTION: EFFECTS OF FAMILIARITY ON THE ABILITY OF OBSERVERS TO DETECT DECEPTION

This study investigated the relationship between an observer's familiarity with the normal, truthful communicative behavior of an individual, and the observer's ability to detect deception on the part of that individual. Specifically, an attempt was made to provide an experimental test of the degree of linearity between familiarity and judgmental accuracy in detecting deception. After exposure to varying amounts of a communicator's normal, truthful behavior (baseline segment), observers made judgments of a communicator's veracity following observation of truthful or lying behavior (test segment). The results indicate a significant deviation from linearity in the relationship between familiarity and judgmental accuracy, and that the relationship is better described by an inverted parabolic curve (quadratic function). The possibility of information overload, the possibility of communicator-specific characteristics which provide clues to deception, and the possibility of observer fatigue were proposed as possible explanations of these results.     

Intuitive and deliberate judgments are based on common principles

A popular distinction in cognitive and social psychology has been between intuitive and deliberate judgments. This juxtaposition has aligned in dual-process theories of reasoning associative, unconscious, effortless, heuristic, and suboptimal processes (assumed to foster intuitive judgments) versus rule-based, conscious, effortful, analytic, and rational processes (assumed to characterize deliberate judgments). In contrast, we provide convergent arguments and evidence for a unified theoretical approach to both intuitive and deliberative judgments. Both are rule-based, and in fact, the very same rules can underlie both intuitive and deliberate judgments. The important open question is that of rule selection, and we propose a 2-step process in which the task itself and the individual's memory constrain the set of applicable rules, whereas the individual's processing potential and the (perceived) ecological rationality of the rule for the task guide the final selection from that set. Deliberate judgments are not generally more accurate than intuitive judgments; in both cases, accuracy depends on the match between rule and environment: the rules' ecological rationality. Heuristics that are less effortful and in which parts of the information are ignored can be more accurate than cognitive strategies that have more information and computation. The proposed framework adumbrates a unified approach that specifies the critical dimensions on which judgmental situations may vary and the environmental conditions under which rules can be expected to be successful.     

Listening, not watching: situational familiarity and the ability to detect deception

In 4 experiments, the authors investigated the influence of situational familiarity with the judgmental context on the process of lie detection. They predicted that high familiarity with a situation leads to a more pronounced use of content cues when making judgments of veracity. Therefore, they expected higher classification accuracy of truths and lies under high familiarity. Under low situational familiarity, they expected that people achieve lower accuracy rates because they use more nonverbal cues for their veracity judgments. In all 4 experiments, participants with high situational familiarity achieved higher accuracy rates in classifying both truthful and deceptive messages than participants with low situational familiarity. Moreover, mediational analyses demonstrated that higher classification accuracy in the high-familiarity condition was associated with more use of verbal content cues and less use of nonverbal cues.     

Contribution of conversation skills to the production of judgmental errors

An impressive body of evidence has accumulated demonstrating that many of the judgmental ‘errors’ or ‘biases’ formerly thought due to purely cognitive shortcomings actually reflect the operation of communication goals and strategies that people rely upon to comprehend and generate meaningful conversation. This study examines the effects of individual differences in conversational skills on the production of biased responses using six judgmental heuristics tasks: base-rate error, conjunction error, dilution effect, underuse of consensus information, primacy effect, and confirmation bias. Clarke's (1975) ‘method of reconstruction’ was used to obtain two measures of conversational sophistication: relevance-seeking and (un)responsiveness. A path analysis predicting biased judgments from the skill variables demonstrates that a combination of these variables, which we term ‘Pragmatic Competence’, is predictive of two independent subsets of the heuristics tasks. Our model provides convergent evidence with other, parametric studies for the proposition that biased social judgments are, at least in part, artifacts of participants' reasonable (and unreasonable!) expectations concerning experimenter cooperativeness. ‘The process of forming an integrated mental model of premises is nothing more than the proper comprehension of discourse: it is required in order to grasp the full impact of what the speaker has to say’ Johnson-Laird (1983, p. 119). © 1998 John Wiley & Sons, Ltd.     

Beyond procedure's content: the role of accessibility experiences and personal uncertainty in procedural justice judgements

Procedural justice concerns play a critical role in economic settings, politics, and other domains of human life. Despite the vast evidence corroborating their relevance, considerably less is known about how procedural justice judgments are formed. Whereas earlier theorizing focused on the systematic integration of content information, the present contribution provides a new perspective on the formation of justice judgments by examining the influence of accessibility experiences. Specifically, we hypothesize that procedural justice judgments may be formed based on the ease or difficulty with which justice-relevant information comes to mind. Three experiments corroborate this prediction in that procedures were evaluated less positively when the retrieval of associated unfair aspects was easy compared to difficult. Presumably this is because when it feels easy (difficult) to retrieve unfair aspects, these are perceived as frequent (infrequent), and hence the procedure as unjust (just). In addition to demonstrating that ease-of-retrieval may influence justice judgments, the studies further revealed that reliance on accessibility experiences is high in conditions of personal certainty. We suggest that this is because personal uncertainty fosters systematic processing of content information, whereas personal certainty may invite less taxing judgmental strategies such as reliance on ease-of-retrieval.     

Dilution and confirmation of probability judgments based on nondiagnostic evidence

Previous research has shown that probability judgments based on a mix of diagnostic and nondiagnostic information are less extreme than judgments based on the diagnostic information alone. Results of the present experiments suggest that this dilution effect holds only under a limited set of conditions. When judgments based on a mix of diagnostic and nondiagnostic information are compared with separately elicited judgments based on the diagnostic information alone, the dilution effect is consistently observed. When judgments based on the diagnostic evidence are revised in light of additional, nondiagnostic evidence, by contrast, the dilution effect is eliminated or even reversed (yielding a confirmation effect) depending on the type of nondiagnostic evidence under evaluation.     

Own-other differences in the realism of some metacognitive judgments

The present study investigated differences in judgments of one's own and others' knowledge (the own-other difference). Consistent with the below-average effect (e.g., Kruger, 1999), our main results showed that the participants gave lower knowledge ratings of their own extent of knowledge than of another person's extent of knowledge (Experiment 1). Furthermore, lower and more realistic judgments were found when the participants judged their own as compared with when judging another person's overall accuracy (frequency judgments) of answering knowledge questions correctly (Experiment 1 and 2). On the basis of these results it is argued that judgmental anchoring may be important also in the context of indirect comparisons, and that previous conclusions of cross-cultural psychology regarding the above-average effect may be oversimplified.     

Motivational determinants of systematic processing: expectancy moderates effects of desired confidence on processing effort

Extending the motivational assumptions of the heuristic-systematic model (Chaiken, Liberman, & Eagly, 1989), the authors hypothesized that a discrepancy between desired and actual judgmental confidence raises processing effort only if the expectancy that processing will increase confidence is high. In Experiment 1, university students expected to review information for upcoming social judgments. Desired confidence was varied through low versus high task importance. To manipulate expectancy, low versus high perceived processing efficacy was induced via feedback. As predicted, high- (as compared to low-) importance participants expressed greater interest in receiving information and selected more information when perceived efficacy was high, and this effect was mediated via a heightened discrepancy between desired and actual confidence. These effects were not obtained under low perceived efficacy. In Experiment 2, students processed a persuasive message. Only high importance conditions were studied; processing efficacy and argument strength were manipulated. As predicted, high- (but not low-) efficacy participants processed the message systematically, as indicated by a different impact of argument strength and by mediational path analyses. It is argued that the precision of social judgment models would benefit from an explicit consideration of processing- and outcome-related expectancy variables. © 1998 John Wiley & Sons, Ltd.     

Non-communicated judgements of,versus feedback on,students’ essays: Is feedback inflation larger for students with a migration background?

When providing feedback, teachers are concerned not only with the simple transmission of information, but also with motivational and interpersonal dynamics. To mitigate these concerns, teachers may inflate feedback by reducing negative or increasing positive content. The resulting difference between initial judgments and feedback may be even more drastic for ethnic minority students: In non-communicated judgments, negative stereotypes may result in more negative judgments, whereas in feedback, concerns about being or appearing prejudiced may inflate feedback towards ethnic minority students. These hypotheses were tested in a sample of 132 German teacher students in a 2 (between subjects: feedback vs. non-communicated judgment)?×?2 (within subjects: target student's migration background: Turkish vs. none) design in which participants read supposed student essays and provided their written impressions to the research team or the supposed student. Findings revealed that teacher students’ feedback was more positive than their non-communicated judgments on a multitude of dimensions. Contrary to expectations, these effects were not stronger when the student had a Turkish migration background. Instead, teacher students rated the essay of the student with a Turkish migration background more favorably both in the judgment and feedback conditions. Our results suggest that teachers adapt their initial judgments when giving feedback to account for interpersonal or motivational dynamics. Moreover, ethnic minority students may be especially likely to receive overly positive feedback. While the motivational/interpersonal dynamics may warrant some inflation in feedback, negative consequences of overly positive feedback, for which ethnic minority students may be especially vulnerable, are discussed.