H1: Participants will gather less evidence (draw fewer beads) in the social condition than in the asocial condition ("JTC in a relative sense").
The study showed that the draws to decision in the social condition were significantly lower than the asocial condition (figure 3a), therefore, social information amplified JTC in a relative sense.


H2: Participants will be more likely to jump to conclusions (making objectively premature decisions) in the social condition than in the asocial condition (“JTC in an absolute sense”).
Results showed that participants were significantly more likely to jump to conclusions in the social condition relative to the asocial control (figure 3b). Thus, the social context amplified the JTC bias.


H3: Participants in the social condition will make their decisions about the majority color of the urn at a point of lower objective certainty than will participants in the asocial condition.
Results showed that participants in the social condition decided at a significantly lower objective certainty.


H4: Participants will be more likely to form inaccurate beliefs about the state of the world in the social condition, relative to the asocial control.
Results showed that participants in the social condition were significantly less accurate.

Further Results and Hypotheses for Social Information Experiment

- What, if anything, is cognitively distinctive about the participants who were susceptible to the faulty intuition that the evidence was worthless or who tended to imitate this heuristic in others?
- What patterns in social information are responsible for the study's main effect?

Questions for Future Research on Social Information and Collective Delusions

The study found a significant effect of condition for all four preregistered hypotheses but no effect of position and no interaction between condition and position. A social context has thus amplified the jumping to conclusions (JTC) bias, increasing the likelihood that people would make costly decisions on the basis of insufficient evidence.

Syracuse University

In order to test the effects of social learning on individual learning the study had participants do a probabilistic reasoning task, the bead task. All participants were told that the urn full of beads either had 40 pink beads and 60 green or 40 green beads and 60 pink beads and their goal was to guess the correct majority. Money would be given to those who guessed correctly in order to incentivize the participants. The participants could choose to look at the colors of 0-15 beads, however the more beads they picked, the more money was subtracted from their $.15 endowment. The control group completed this task without any given social information of the other participants’ actions. The experimental group was provided social information (irrelevant to the actual answer) that only told them how many beads had been drawn by the previous participant, the colors of those beads were not included in the information.

Experiment Set-Up for Study of How Social Information Amplifies the Tendency to Gather Insufficient Data

Sulik, J., Efferson, C., &amp; McKay, R. (2021). Collectively jumping to conclusions: social information amplifies the tendency to gather insufficient data. https://doi.org/10.31234/osf.io/vpnxh 

Learn Before

Related

Learn After