Investigating people's uncertainty without changing it: non-numeric measurement and conceptualization of psychological uncertainty
Five experiments compared the use of verbal and numeric methods of assessing people's uncertainty and investigated the hypothesis that numeric measures of uncertainty can prompt people to engage in a deliberate and rule-based mode of thinking that is atypical for many situations. In Experiments 1 through 4, participants read scenarios about situations involving uncertainty and provided decisions or judgments in reaction to the described situations. In Experiment 5, participants encountered gambling situations and their betting behaviors were recorded. For each of the five experiments, one group of participants provided numeric estimates of uncertainty before providing judgments, decisions, or behaviors (numeric-first group), another group provided verbal estimates of uncertainty before providing judgments, decisions, or behaviors (verbal-first group), and another group provided no uncertainty estimates before providing judgments, decisions, or behaviors (control-group). If numeric measures of uncertainty prompt people to engage in an atypically deliberate and rule-based form of processing, the responses made in the wake of such processing might be affected. Accordingly, it was hypothesized that the judgments, decisions, and behaviors of participants in numeric-first groups would differ from those in control and verbal-first groups. The experiments did not yield the expected support for the prompting hypothesis. Experiments 3 and 4, however, provided initial demonstrations of what was named the alternative-outcomes effect. Verbal measures of uncertainty revealed that the perceived likelihood of a target outcome can be influenced by normatively equivalent manipulations to the distributions of alternative outcomes. The possible importance of this alternative-outcomes effect is discussed. Also, the influence of base-rate information on people's uncertainty is discussed in conjunction with Experiments 1 and 2, and the effects of perceived control on uncertainty is discussed in conjunction with Experiment 5.