Peter Juslin

Exemplar effects in categorization and multiple-cue judgment.

Categorization and multiple-cue judgment are similar tasks, but the influential models in the two areas are different in terms of the computations, processes, and neural substrates that they imply. In categorization, exemplar memory is often emphasized, whereas multiple-cue judgment generally is interpreted in terms of integration of cues that have been abstracted in training. In 3 experiments the authors investigated whether these conclusions derive from genuine differences in the processes or are accidental to the different research methods. The results revealed large individual differences and a shift from exemplar memory to cue abstraction when the criterion is changed from a binary to a continuous variable, especially for a probabilistic criterion. People appear to switch between qualitatively distinct processes in the 2 tasks.

Thurstonian and Brunswikian origins of uncertainty in judgment: A sampling model of confidence in sensory discrimination

As a preliminary step towards the presentation of a model of confidence in sensory discrimination, the authors propose a distinction between 2 different origins of uncertainty named after 2 of the great probabilists in the history of psychology, L.L. Thurstone and Egon Brunswik. The authors review data that suggest that there are empirical as well as conceptual differences between the 2 modes of uncertainty and thus that separate models of confidence are needed in tasks dominated by Thurstonian and Brunswikian uncertainty. The article presents a computational model for 1 class of tasks dominated by Thurstonian uncertainty: sensory discrimination with pair comparisons. The sensory sampling model predicts decisions, confidence assessments, and the complex pattern of response times in simple psychophysical discrimination tasks (J.V. Baranski and W.M. Petrusic, 1994). The model also accounts for the disposition towards underconfidence often observed in sensory discrimination with pair comparisons.

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 Naïve intuitive statistician : A naïve sampling model of intuitive confidence intervals

The perspective of the naïve intuitive statistician is outlined and applied to explain overconfidence when people produce intuitive confidence intervals and why this format leads to more overconfidence than other formally equivalent formats. The naïve sampling model implies that people accurately describe the sample information they have but are naïve in the sense that they uncritically take sample properties as estimates of population properties. A review demonstrates that the naïve sampling model accounts for the robust and important findings in previous research as well as provides novel predictions that are confirmed, including a way to minimize the overconfidence with interval production. The authors discuss the naïve sampling model as a representative of models inspired by the naïve intuitive statistician.

An explanation of the hard-easy effect in studies of realism of confidence in one's general knowledge

Abstract Recent ecological approaches to realism of confidence in general knowledge (Gigerenzer, Hoffrage & Kleinbolting, 1991; Juslin, in press) argue that people are well-calibrated to their natural environments. Both the overconfidence phenomenon and the hard-easy effect are explained as consequences of informal experimenter-guided selection of almanac items, selection that changes the validity of the cues used by the subjects for selection of answers to the items. The paper presents the ecological approach and reports an experiment showing that: (1) when the objects of judgement are selected randomly from a natural environment, people are well-calibrated; (2) when more and less difficult item samples are created by selecting items with more and less familiar contents, i.e. in a way that does not affect the validity of the cues, no hard-easy effect is observed, and people are well-calibrated both for hard and easy item samples. These results, predicted by the ecological approach, provide further suppor...

Format dependence in subjective probability calibration

Empirical data from 2 experiments confirmed the format dependence predicted by the combined error model (P. Juslin, H. Olsson, & M. Bjorkman, 1997). Format dependence refers to the simultaneous observation of over/underconfidence in judgment for the same

Brunswikian and Thurstonian Origins of Bias in Probability Assessment: On the Interpretation of Stochastic Components of Judgment

Uppsala University, SwedenABSTRACTIn this paper the Brunswikian framework provided by the theory of ProbabilisticMental Models (PMM), and other theoretical stances inspired by probabilisticfunctionalism, is combined with the Thurstonian notion of a stochastic com-ponent of judgment. We review data from 25 tasks with representative selection ofitems collected in our laboratory. Over/underconfidence is close to zero in mostdomains, but there is a moderate hard–easy e•ect across task domains that isinconsistent with the original assumptions of the Brunswikian framework. Thebinomial model modifies PMM-theory by allowing for sampling error in theprocess of learning the ecological probabilities and the response-error model takeserror in the process of overt probability assessment into account. Both modelspredict a moderate hard–easy e•ect across task environments that di•er indi†culty or predictability, but it is also demonstrated that the two interpretationsof random error lead to di•erent predictions. The response error model predictsformat dependence, with more overconfidence in full-range than in half-rangeassessment, and the phenomenon is illustrated with empirical data. It is proposedthat a model that combines the Brunswikian framework with both sampling errorand response error captures many of the important phenomena in the calibrationliterature. For illustrative purposes, a combined model with four parameters isfitted to empirical data suggesting good fit. #1997 by John Wiley & Sons, Ltd.

Information integration in multiple cue judgment: a division of labor hypothesis.

There is considerable evidence that judgment is constrained to additive integration of information. The authors propose an explanation of why serial and additive cognitive integration can produce accurate multiple cue judgment both in additive and non-additive environments in terms of an adaptive division of labor between multiple representations. It is hypothesized that, whereas the additive, independent linear effect of each cue can be explicitly abstracted and integrated by a serial, additive judgment process, a variety of sophisticated task properties, like non-additive cue combination, non-linear relations, and inter-cue correlation, are carried implicitly by exemplar memory. Three experiments investigating the effect of additive versus non-additive cue combination verify the predicted shift in cognitive representations as a function of the underlying combination rule.

Cue abstraction and exemplar memory in categorization

In this article, the authors compare 3 generic models of the cognitive processes in a categorization task. The cue abstraction model implies abstraction in training of explicit cue-criterion relations that are mentally integrated to form a judgment, the lexicographic heuristic uses only the most valid cue, and the exemplar-based model relies on retrieval of exemplars. The results from 2 experiments showed that, in lieu of the lexicographic heuristic, most participants spontaneously integrate cues. In contrast to single-system views, exemplar memory appeared to dominate when the feedback was poor, but when the feedback was rich enough to allow the participants to discern the task structure, it was exploited for abstraction of explicit cue-criterion relations.

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.