The optimist within? Selective sampling and self-deception

- We combined paradigms from the “optimism bias” and “crowd within” literatures.
- Participants provided two different estimates for each of a series of questions.
- First and second estimates for neutral items did not differ systematically.
- Second estimates for undesirable items were more optimistic than first estimates.
- Participants sample selectively from within when estimating undesirable events.

The nature and existence of self-deception is controversial. On a classic conception, self-deceived individuals carry two conflicting representations of reality. Proponents of an alternative, deflationary account dispute this, arguing that putative cases of self-deception simply reflect distorted information processing. To investigate these alternatives, we adapted a paradigm from the “crowd-within” literature. Participants provided two different estimates for each of a series of incentivized questions. Half of the questions were neutral in content, while half referred to undesirable future events. Whereas the first and second estimates for neutral questions did not differ systematically, second estimates for undesirable questions were more optimistic than first estimates. This result suggests that participants were sampling selectively from an internal probability distribution when providing estimates for undesirable events, implying they had access to a less rosy representation of their future prospects than their individual estimates conveyed. In short, self-deception is real.