This paper empirically analyses two types of preferences over the timing of resolution of uncertainty: preferences between early and late resolution and preferences between one-shot and gradual resolution of lotteries under risk and ambiguity. In an online experiment with students, we find significant differences between treatments: under risk, a majority of participants show a strict preference against gradual resolution of uncertainty, for low, medium and high ex-ante probabilities of receiving the prize of the lottery. Under ambiguity, most participants show a preference for gradual resolution of uncertainty for lotteries with a low-likelihood of winning, and an aversion towards it for medium and high likelihoods. Additionally, in both treatments we find subjects show strict preferences more frequently in the one-shot vs. gradual resolution dimension than in the early vs. late resolution dimension. Results from the experiment contribute to the literature about the empirical validity of ambiguity models, as different models prescribe different preferences over the timing of the resolution of uncertainty.

In this paper, I analyse how modal decisions within social groups influence decision-making under uncertainty. Specifically, I study how these modal decision are more often taken by members of the group as uncertainty increases. In a theoretical model, in which distancing from group's modal decision is penalised as a regret cost, I find that aligning with the in-group's modal decision happens more often the higher is the identification with the group and it is also more frequent under ambiguity than under risk. I design and run an online experiment with a sample of US citizens in which social groups are formed through political affiliation. Results show that higher self-identification with the group leads to higher alignment with the group's modal decision, under riks. Ambiguity generates an alignment with modal decisions of subjects with low- or medium- self-identification similar to the one of high-identification subjects in the risk treatment.