Guessing imagined and live chance events: Adults behave like children with live events

Abstract

An established finding is that adults prefer to guess before rather than after a chance event has happened. This is interpreted in terms of aversion to guessing when relatively incompetent: After throwing, the fall could be known. Adults (N=71, mean age 18;11, N=28, mean age 48;0) showed this preference with imagined die‐throwing as in the published studies. With live die‐throwing, children (N=64, aged 6 and 8 years; N=50, aged 5 and 6 years) and 15‐year‐olds (N=93, 46) showed the opposite preference, as did 17 adults. Seventeen‐year‐olds (N=82) were more likely to prefer to guess after throwing with live rather than imagined die‐throwing. Reliance on imagined situations in the literature on decision‐making under uncertainty ignores the possibility that adults imagine inaccurately how they would really feel: After a real die has been thrown, adults, like children, may feel there is less ambiguity about the outcome.

Agency Affects Adults', but not Children's, Guessing Preferences in a Game of Chance

“…Table 2 demonstrates children's guessing preferences both when they picked a pen and drew the circle, and when these were done by the experimenter. Unlike the adult data, there was no difference in guessing preferences according to who controlled the outcome, In line with the findings reported in Beck et al (2010) and Robinson et al (2009;also, McColgan et al, 2010), children preferred to guess under conditions of epistemic rather than physical uncertainty, and this effect was no different whether the child themselves, or the experimenter, picked the pen and drew the circle.…”

“…For adults, we observed precisely that result. Children, however, always preferred to guess under conditions of epistemic uncertainty, consistent with previous results and theory McColgan et al, 2010;Robinson et al, 2006Robinson et al, , 2009. By setting up a situation in which adults' and children's guessing preferences disassociate, we have provided the first evidence that their guessing preferences are driven by different underlying cognitive mechanisms.…”

“…Consequently, if there is no difference between conditions in children's guessing preferences, then a moderated competence hypothesis would appear to be a less valid hypothesis for explaining children's treatment of uncertainty than it is for adults. Such a result would demonstrate that adults' and children's guessing preferences obtain for different reasons, thus answering the final question posed in Robinson et al (2009).…”

“…We predicted that when the outcome was controlled by the experimenter, adults would prefer to guess under epistemic uncertainty (in line with Robinson et al, 2009), because no blame could be attributed to them for a false guess, whilst they would prefer to guess under physical uncertainty when they controlled the outcome themselves (due to the perceived potential attribution of blame).…”

“…Having observed evidence consistent with the effect of relative (in)competence on adults' guessing preferences in a situation in which they themselves controlled the outcome, Study 2 is the first test of whether feelings of relative (in)competence might affect children's guessing preferences. Robinson et al (2009) detailed that, in order to be sensitive to feelings of relative (in)competence, children must be able to engage in counterfactual thinking, and compare an evaluation of the counterfactual state with the experienced reality. Robinson et al thus conjectured that, as these same cognitive pre-requisites are required for the experience of regret, if children are too young to experience regret, they will also be too young to be sensitive to feelings of relative (in)competence.…”