Article #14
Newsletter list 


Previous Article 

Next Article 


Research in Uppsala

Peter Juslin
Uppsala, Sweden

Our current research in Uppsala can be summarized in four points: First, we follow up on previous research. Here are two examples. We (Anders Winman, Henrik Olsson and I) have recently finished a first version of a paper that contains a "meta-analysis" of all published or otherwise available data on realism of confidence in studies with two-alternative general knowledge items. The study investigates a) if representative vs selected item samples has an effect on over/underconfidence, b) if there is evidence for a cognitive overconfidence bias when the items have been representatively selected, and c) if there remains a substantial hard-easy effect in these data once that we control for selection effects and more trivial statistical effects. We (mainly Anders Winman and I) are also starting to analyze a data set that provides further comparisons between Thurstonian and Brunswikian error in judgment, e.g., with respect to additivity, over/underconfidence in interval estimation, and "ambiguity avoidance". Henrik Olsson is continuing the work on the sensory sampling model, among other things, applying it to the case of perceptual bias.

Second, we (mainly Anders Winman and I) have performed two experiments that aim to investigate the nature of the cognitive representations that underlie judgment in MCPL. We are specifically interested in finding ways to figure out when judgments are mediated by rule-based processing vs memory-based processing. This analysis is still in progress.

Third, we (Magnus Persson and I) are working on an exemplar-based model of subjective probability assessment, entitled PROBEX. The model is intended to describe memory-based judgments and predicts decisions, point-estimates, subjective probabilities, and response times. We are currently finnishing the computer-simulations and the data collection. Again, this is work on the way right now.

Finally, we (mainly Pia Wennerholm and I) are performing studies on base-rate use in judgment, both in more simple categorization designs and more ecologically relevant contexts.

As is evident much of this research is currently in progress, so the fertility of many of our efforts is still uncertain. This is only as it should be, of course."

Contact Peter Juslin

Previous Article                Next Article

Newsletter list