Tracking altercentric and egocentric bias in a continuous False Belief task

• In Event: Poster Session 07
  In Poster Session: PS 07 Topic - Social Cognition

Abstract

Social cognition is subject to specific biases. Egocentric bias refers to the phenomenon that subjects are slowed down and led into error by their own perspective, even when they are explicitly asked to focus on others’ perspectives (e.g., Samson et al., 2010). More surprisingly, recent studies provided evidence for an altercentric bias as well: our own judgments and behavior are modulated by how we think other agents perceive the world, indicating that we implicitly represent their perspectives, even when those are irrelevant or interfere with our own task (e.g., Southgate, 2020). In our upcoming study, we plan to investigate both types of biases within one task format that combines the methods of previous studies: The so-called Sandbox task measures egocentric bias in a modified standard false belief scenario with continuous rather than discrete locations: when asked where an agent believes an object is, subjects deviate from the correct answer somewhat in the direction of where they themselves know the object to be – and the degree of deviation reflects the magnitude of the egocentric bias (Sommerville et al., 2013). Mouse tracking measures have been used previously to document altercentric bias (van der Wel et al., 2014): when subjects are asked to move the mouse cursor to the location of a target object, they take a little detour on their way to the right answer when another agent in the scenario has a deviant belief; the area under this detour thus indicates whether/to what degree participants engage in altercentric bias (see Figure 1). This is a proof-of-concept study which aims to lay the foundations for mapping the developmental trajectories of these biases.
In a new hybrid task, these measures will be combined in the following ways: In explicit conditions that tap egocentric bias, participants are asked to indicate where the protagonist believes an object is. If participants are affected by their own perspective/knowledge when judging others’ behaviors, we expect their answer to be somewhere around Location 1 (correct answer) but biased toward Location 2 (the object’s real location). And also, we expect them to draw a convex bow detouring toward Location 2 when moving the cursor to mark their answers.
Altercentric bias will be measured by the implicit version of the same task, in which participants are asked to indicate where the object really is. In the experimental trials, agents’ and participants’ perspectives differ from each other. If participants engage in implicit perspective taking and represent the agent’s mistaken -although irrelevant- perspective, we would expect their judgments to become subject to altercentric interference such that their answers deviate from the object’s real location (Location 2) toward the agent’s belief location (Location 1). And also, we expect them to draw a convex bow detouring toward Location 1 when moving the cursor to mark their answers.
In a mixed design, 94 German adults (age>=18) will be tested online on either the explicit or the implicit version, with 2 types of trial (experimental and control) per version administered within subjects.

Authors

• Feride Nur Haskaraca Kizilay, Georg-August-Universität Göttingen (GAU)
  Presenting Author

• Marina Proft, Georg-August-Universität Göttingen
  Non-Presenting Author

• Hannes Rakoczy, University of Göttingen
  Non-Presenting Author