• People
  • Senior people
    • Michael Dougherty
    • Rick Thomas
  • Post Doctoral Students
  • Graduate Students
  • Collaborators
  • Alumni
• Research
  • Hypothesis Generation
  • Probability Judgment
  • Hypothesis Testing
  • Risk
  • Memory Research
  • Attention And Mood
  • Non-Metric Regression
• Publications
• Phun Stuff
• About Us

 

Probability and Confidence Judgment

The nature of how people make judgments of probability or confidence has been of interest to social scientists for decades. Within the judgment and decision making literature, probability judgments are of interest because they form one piece of the Expected Utility theory model and because they are assumed to form the basis of peoples risk perceptions. Therefore, of particular importance for judgment and decision researchers is understanding the degree to which judgments of probability depart from some 'normative' or 'rational' standard. In most JDM applications, the researcher is interested assessing the accuracy of participants judgments in terms of gauging the occurance of an event that is external to the decision maker (e.g., what is the probability that Tom has cancer?, or what is the probability that XYZ stock will increase in the first quarter?). In the broader psychological literature, confidence judgments have often been used as a means of assessing an internal event, such as the accuracy of a memory, or the prediction of the probabilty that one will successfully retrieve an item for an upcoming exam.

Our research blurs the line between assessments of external events and assessments of internal events. We assume that the memory system provides the input to the judgment process, regardless of whether it is a judgment about something external or internal to the participant. Thus, errors, biases, or constraints in the memory system can cascade into error and biases in judgment. Several studies attest to the dependence of judgment on attention and memory processes. Here are a few examples:

• Judgment accuracy depends on how well information is encoded in memory.
  • Overconfidence decreases as encoding quality increases (Dougherty, 2001).
  • Divided attention at encoding leads to increased Subadditivity (Dougherty, Atkins, & Thomas, in prep).
• Judgment is capacity limited.
  • Degree of subadditivity is negatively correlated with individual differences in working memory (Dougherty & Hunter, 2003; Sprenger & Dougherty, 2006).
  • Divided attention during judgment leads to an increase in judged (Dougherty et al., submitted).
• Judgments of probability are succeptible to the buildup of proactive interference.
  • Relative accuracy always decreases with the build up of proactive interference. Degree of subadditivity is affected by 'strength' of the interfering items (Dougherty & Sprenger, 2006).