Advice Taking in Groups

• Applicant: Professor Dr. Stefan Schulz-Hardt, since 4/2022
• Subject Area: Social Psychology, Industrial and Organisational Psychology
• Term: since 2016
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 314153918

Groups make less use of an individual's advice than individuals do when forming judgements. To better understand this behavior, we developed a model in the first funding period that accurately predicts the use of advice as a function of group size, provided it is known how individuals use the same advice. The model assumes that groups have a good understanding of how the value of independent advice decreases with group size, assuming that group members in turn bring their independent opinions to the discussion of a matter. Since the assumption of independent individual opinions within groups is not always tenable, our model assumes that groups do not recognize possible dependencies, but instead always behave as if members were independent. Finally, the model assumes that the resistance to advice typical for individuals persists unchanged in groups. Using this simple model, we were able to accurately predict the weight of advice for groups of different sizes if the group members made independent individual judgements prior to the group discussion (independent groups). However, without such individual judgements (dependent groups) our model underestimated how much groups heeded the advice. In the first part of the follow-up proposal we aim to test an explanation for the different fit of our model for dependent and independent groups. This explanation states that advice taking in groups is based on how many members actually participate in the discussion (functional group size) and not on the mere number of group members (actual group size). Accordingly, we want to test whether our model can predict the use of advice in groups in a context-independent way when we calculate its predictions based on the functional group size. So far, we have only investigated situations with a single person as advisor. In reality, however, it is often the case that advice is based on the opinions of several people, either by aggregating independent opinions (as in the case of online product evaluations) or by discussion within groups (for example, when an expert panel makes a recommendation). In the second part of the proposal we therefore want to extend our model to situations where groups receive advice from several people. Specifically, for both types of advice, we want to examine whether an extension of our model can accurately predict how much groups (and individuals) use advice that incorporates the opinions of several people.

• DFG Programme: Research Grants
• Ehemaliger Antragsteller: Dr. Thomas Schultze-Gerlach, until 4/2022