January 18, 2018

From Journal of Personality and Social Psychology:

Across 6 studies we investigated the development of overconfidence among beginners. In 4 of the studies, participants completed multicue probabilistic learning tasks (e.g., learning to diagnose "zombie diseases" from physical symptoms). Although beginners did not start out overconfident in their judgments, they rapidly surged to a "beginner's bubble" of overconfidence. This bubble was traced to exuberant and error-filled theorizing about how to approach the task formed after just a few learning experiences. Later trials challenged and refined those theories, leading to a temporary leveling off of confidence while performance incrementally improved, although confidence began to rise again after this pause. In 2 additional studies we found a real-world echo of this pattern of overconfidence across the life course. Self-ratings of financial literacy surged among young adults, then leveled off among older respondents until late adulthood, where it begins to rise again, with actual financial knowledge all the while rising more slowly, consistently, and incrementally throughout adulthood. Hence, when it comes to overconfident judgment, a little learning does appear to be a dangerous thing. Although beginners start with humble self-perceptions, with just a little experience their confidence races ahead of their actual performance.

From Intelligence:

The present experiment (N = 390) examined how people adjust their judgment after they learn that crucial information on which their initial evaluation was based is incorrect. In line with our expectations, the results showed that people generally do adjust their attitudes, but the degree to which they correct their assessment depends on their cognitive ability. In particular, individuals with lower levels of cognitive ability adjusted their attitudes to a lesser extent than individuals with higher levels of cognitive ability. Moreover, for those with lower levels of cognitive ability, even after the explicit disconfirmation of the false information, adjusted attitudes remained biased and significantly different from the attitudes of the control group who was never exposed to the incorrect information. In contrast, the adjusted attitudes of those with higher levels of cognitive ability were similar to those of the control group. Controlling for need for closure and right-wing authoritarianism did not influence the relationship between cognitive ability and attitude adjustment. The present results indicate that, even in optimal circumstances, the initial influence of incorrect information cannot simply be undone by pointing out that this information was incorrect, especially in people with relatively lower cognitive ability.

From Thinking & Reasoning:

Initial acts of self-control can reduce effort and performance on subsequent tasks – a phenomenon known as ego depletion. Ego depletion is thought to undermine the capacity or willingness to engage executive control, an important determinant of success for many tasks. We examined whether ego depletion improves performance on a task that favours less executive control: insight problem solving. In two experiments, participants completed an ego-depletion manipulation or a non-depleting control condition followed by an insight problem-solving task (i.e., matchstick arithmetic). Participants in the depleting condition demonstrated greater insight problem-solving accuracy than those in the non-depleting control condition. Priming theories of willpower did not impact these results. Although ego depletion is widely regarded as a “state of impairment”, attendant decreases in executive control may foster insightful thinking.

From Google Working Paper:

The game of chess is the most widely-studied domain in the history of artificial intelligence. The strongest programs are based on a combination of sophisticated search techniques, domain-specific adaptations, and handcrafted evaluation functions that have been refined by human experts over several decades. In contrast, the AlphaGo Zero program recently achieved superhuman performance in the game of Go, by tabula rasa reinforcement learning from games of self-play. In this paper, we generalise this approach into a single AlphaZero algorithm that can achieve, tabula rasa, superhuman performance in many challenging domains. Starting from random play, and given no domain knowledge except the game rules, AlphaZero achieved within 24 hours a superhuman level of play in the games of chess and shogi (Japanese chess) as well as Go, and convincingly defeated a world-champion program in each case.

From Law and Human Behavior:

Research examining how truth tellers' and liars' verbal behavior is attenuated as a function of delay is largely absent from the literature, despite its important applied value. We examined this factor across 2 studies in which we examined the effects of a hypothetical delay (Experiment 1) or actual delay (Experiment 2) on liars' accounts. In Experiment 1-an insurance claim interview setting-claimants either genuinely experienced a (staged) loss of a tablet device (n = 40) or pretended to have experienced the same loss (n = 40). Truth tellers were interviewed either immediately after the loss (n = 20) or 3 weeks after the loss (n = 20), whereas liars had to either pretend the loss occurred either immediately before (n = 20) or 3 weeks before (n = 20) the interview (i.e., hypothetical delay for liars). In Experiment 2-a Human Intelligence gathering setting-sources had to either lie (n = 50) or tell the truth (n = 50) about a secret video they had seen concerning the placing of a spy device. Half of the truth tellers and liars where interviewed immediately after watching the video (n = 50), and half where interviewed 3-weeks later (n = 50; i.e., real delay for liars). Across both experiments, truth tellers interviewed after a delay reported fewer details than truth tellers interviewed immediately after the to-be-remembered event. In both studies, liars failed to simulate this pattern of forgetting and reported similar amounts of detail when interviewed without or after a delay, demonstrating a stability bias in reporting.