Table 1. Experimental Setup. The experiment consisted of 12 sessions (S) each composed of 12 subjects. In each session, the 12 subjects had to answer two questions (Q) in the no information, two in the aggregate and two in the control condition (see main text), for a total of six questions. The order of the questions was randomised across sessions. After each of the five rounds subjects were asked the same question again and could revise their answers depending on the information available to them. In the table, columns indicate question number and rows – information regime. Each cell lists the sessions when a given question was asked for a particular information regime.

	Q1	Q2	Q3	Q4	Q5	Q6
no info	S1,S4, S7,S10	S2,S5, S8,S11	S3,S6, S9,S12	S1,S4, S7,S10	S2,S5, S8,S11	S3,S6, S9,S12
agregate info	S3,S6, S9,S12	S1,S4, S7,S10	S2,S5, S8,S11	S3,S6, S9,S12	S1,S4, S7,S10	S2,S5, S8,S11
full info	S2,S5, S8,S11	S3,S6, S9,S12	S1,S4, S7,S10	S2,S5, S8,S11	S3,S6, S9,S12	S1,S4, S7,S10