‘Why did that person just run that red light? They obviously don’t know how to drive.’
We hear it all the time, the tendency to attribute malice or incompetence to another individual or group, when if it was us doing the action it would be merely an accident: ‘I just didn’t see it.’ Welcome to the second edition of Cognitive Bias Wednesday. While there are many reasons for this tendency, a lot of them stem from a suite of cognitive bias known as Attribution Errors, with the Fundamental Attribution Error (FAE) at their root. Simply put it is the tendency for people to emphasise internal decisions and characteristics for other’s negative actions, while emphasising external factors for their own negative actions. FAE pops its head up in a wide variety of situations, and we probably unconsciously express it every day, it is one of the most powerful decision rationalisation biases.
One classic study of the FAE looked at drinking rates amongst adolescent males, and took two observations: firstly, how much an individual drank, and secondly whether they thought that their peers drank more, the same, or less than them. 1 While actual drinking rates across the group averaged similarly, the attribution of drinking rates amongst the peers was strongly externally inflated. As seen in the title of the study ‘I drink … but not as much as other guys.’ While not attributing incompetence or malice, the negative perception of drinking rates is externally magnified and internally denied. This is despite the drinking rates remaining relatively steady across the cohort. We have the tendency to attribute our own negative characteristics externally, and attribute other’s negative characteristics to their internal space.
Furthermore this is only exacerbated when it is brought into a social setting. While the nature of the FAE is powerful on an individual level it is stronger again amongst groups. The expanded bias, creatively named Group Attribution Error, sees the attributes of the out-group as being defined by individual members of that group. We met this bias briefly in the post a couple of weeks ago on Cyclists vs Motorists and Intergroup biases. This is further expanded again with Pettigrew’s, again creatively named, Ultimate Attribution Error (one must wonder where to go after this). While FAE and GAE look at the ascription to external and out-groups primarily and discard most internal and in-group data, Ultimate Attribution Error seeks to not only explain the demonisation of out-group negative actions, but explain the dismissal of out-group positive behaviours. Interestingly many of the studies that support Pettigrew’s Ultimate Attribution Error look at religio-cultural groups as their case studies, such as the study by Taylor and Jaggi (1974), or later studies on FAE/UAE and suicide bombing (Altran, 2003).
Excursus: One brief and curious aside is that according to one study Protestants appear to be more internally focused, lower rates of FAE/GAE, in comparison with Catholics who are generally externally focused, with higher rates of FAE/GAE. 2 The authors theorise that this is due to an innate greater emphasis on the soul within Protestantism. I will have to look more into their article, and perhaps post on it later.
Excursuses aside, how do these attribution errors affect day to day research and study? One of the ways I think they powerfully affect good academic research and debating is when it comes to the assignment of scholarly labels within academia. I sometimes have students come to me asking if I can point them towards material that is ‘more liberal’ (in the theological sense). Now while I applaud students for wanting to seek alternative views to their own, the level of out-group attribution of ‘liberalism’ commonly leads to a flimsy disagreement with the argument at hand. Commonly it goes like ‘I disagree with this argument because its a liberal argument, and therefore…’ Conversely it works in the opposite fashion ‘I agree with this [flimsy] argument, because we are part of the same group.’ A similar bias is found in several recent articles on the religion and science interface. The argument there commonly goes ‘Religion introduces bias, therefore no confessionally religious people can debate this topic.’ The attribution of innate bias to an out-group, in the same fashion that incompetence is attributed to an observed poor driver, is at play here.
Being aware of our tendency to attribute negative internal characteristics to an out-group participant should help us assess things better in two ways. Firstly it should help us to assess arguments and evidence on the grounds that they are presented, not on the group that they are presented from. In short play the game not the person or group. Stick to the argument and evidence that is set forth and assess it on those grounds, whether you agree or disagree with the person or group who is promulgating it. Secondly, it should help us see blind spots within our own research and work. If we are constantly assessing others based on the same qualities, then we are more likely to be critical with our own research based on the arguments and evidence, rather than letting it float on in-group support.
Attribution errors can be extremely hard to overcome, but knowing about them certainly helps. Hope you have enjoyed this Cognitive Bias Wednesday, as usual weigh in below on the comments!
- Segrist, Dan J., Kevin J. Corcoran, Mary Kay Jordan-Fleming, and Paul Rose. “Yeah, I Drink … but Not as Much as Other Guys: The Majority Fallacy among Male Adolescents.” North American Journal of Psychology 9, no. 2 (June 1, 2007): 307. ↩
- Li, Yexin Jessica, Kathryn A. Johnson, Adam B. Cohen, Melissa J. Williams, Eric D. Knowles, and Zhansheng Chen. “Fundamental(ist) Attribution Error: Protestants Are Dispositionally Focused.” Journal of Personality and Social Psychology 102, no. 2 (February 2012): 281–90. doi:10.1037/a0026294. ↩