Overestimating the Degree to Which Attitudes are Shared between Self and Ingroup versus Outgroup Members

In order for humans to thrive in social interactions, there is a demand in understanding how other people think or feel so that we could respond to them accordingly. Research in the past has shown that people use their own mental states as a reference point to understand others. It would be particularly advantageous for us to use this reference point when there is an overlap between how we and others perceive the world.

Researchers have demonstrated that humans tend to think that people who come from the same social backgrounds think, feel, and behave in a similar fashion.

Unfortunately, we often underestimate this perceived overlap between ourselves and people from our own group. A research conducted by Brandon Woo and Jason Mitchell from Harvard University demonstrated this distorted reality. They asked two main research questions: (1) to what extent are people correct in assuming their attitudes are more like those of ingroup members (people who are like them) versus outgroup members (people who are different from them)? and (2) if people are incorrect about this perceived overlap between themselves and others, are they incorrect about how different they are from ingroup members, from outgroup members, or both?

To test these research questions, Woo and Mitchell introduced their participants to two pictures: one who identified as conservative and one who identified as liberal. After that, participants received 100 apolitical attitude statements (e.g., enjoy exercising), and were asked to rate how well these statements applied to themselves, the person who identified as liberal, and the person who identified as conservative. Apolitical statements were chosen because they were no stereotypes that link these statements with any political affiliations.

For each of these 100 statements, Woo and Mitchell calculated a difference score between an estimated difference (the difference between a self-rating and rating for a target) and an actual difference (the difference between a self-rating and the mean rating of all participants in a target’s group).

From their theoretical view, positive score indicates that participants are overestimating how different they are from the target; a negative score indicates participants are underestimating how different they are from a target.

The results of their study were consistent with what they had hypothesized. In general, participants more strongly underestimated the difference between themselves and an ingroup target. For instance, participants who identified as conservative tended to think that the conservative target was very similar to themselves. The result flipped when the target was a liberal: participants tended to think that the liberal target was significantly different than themselves.

Overall, although people use mental states as a model to predict how other people think or feel, they are incorrect, unfortunately. The notion that our own mental states could serve as a good reference to infer other people’s thoughts and emotions is not as reliable as we thought it was.

By: M. Fazuan (Faz) Abdul Karim. Faz is a 2nd year doctoral student in the Applied Social and Organizational Psychology program at Indiana University-Purdue University Indianapolis (IUPUI), working with Leslie Ashburn-Nardo and Evava Pietri. His program of research focuses on two diversity-related themes: the first line of research examines compensatory strategies used by members of stigmatized racial groups to deflect discrimination and the second line of research investigates subtler forms of biases for individuals who are considered non-prototypical of their ingroup.