Some interesting new research by Brendan Nyhan and Jason Reifler from Dartmouth examines attempts to change people’s strongly-held pre-existing beliefs. This kind of research is very important, especially for progressive organisations and causes, and it ties into what my previous blog post about evidence-based campaigning.

The research looks at (mis)perceptions of three issues: the war in Iraq, job growth under Obama and global warming. In particular, they use graphs to display complex information, such as a graph showing the number of jobs created under Obama, and compare it to a paragraph of text explaining the same information.

From the abstract:

People often resist information that contradicts their preexisting beliefs. This disconfirmation bias is a particular problem in the context of political misperceptions, which are widespread and frequently difficult to correct. In this paper, we examine two possible explanations of the prevalence of misinformation. First, people tend to resist unwelcome information because it is threatening to their worldview or self-concept. Drawing from social psychology research, we therefore test whether affirming individuals’ self-worth and thereby buttressing them against this threat can make them more willing to acknowledge uncomfortable facts. Second, corrective information is often presented in an ineffective manner. We thus also examine whether graphical corrections may be more effective than text at reducing counter-arguing by individuals inclined to resist counter-attitudinal information.

There’s a lot to read through in this article, and it is definitely worth reading in detail (and the PDF is free to access).

The main take aways from this research is that there is a right way and wrong way to present complex information that contradicts peoples’ deep-held beliefs, and that graphs can help reduce misconceptions. (Of course, graphs are not a silver bullet.) The research is also clear that just using text to present information that contradict’s someone’s deeply held beliefs just doesn’t work:

When people encounter dissonant information, it is threatening to their self-concept, which they seek to maintain by either dissonance reduction or other strategies for affirming their self-worth… In particular, they tend to interpret ambiguous or mixed information in line with their preexisting views and to resist or reject counter-attitudinal information. For instance, Lord, Ross, and Lepper (1979) found that individuals who were presented with balanced information about the effectiveness of capital punishment rated the counter-attitudinal information to be less convincing and of lower quality than pro-attitudinal information.

In addition to using graphical information, Nyhan and Reifler also test using “affirmation” tests, where they ask respondents to recall pleasant experiences or thoughts about themselves before presenting them with “dissonant” information. This can reduce misconceptions. Affirmation combined with graphical representation has the greatest effect.

For progressive organisations, this kind of research should provoke some thoughts about how we present information. For example:

• Can we simplify our message so we reduce the possibility we will be misunderstood?
• Can we present any of the information we’re trying to get across in a different way? For example, with graphics, or graphs?
• Have we written or presented the information in a way that’s mindful of our audience?
• When presenting confronting information, can we reduce the confrontation by “affirming” the reader?

Again, it’s worth pointing out that this research doesn’t look at how to change minds but rather how to correct misperceptions. This research shows that by presenting information graphically and using affirmation, even people who are predisposed to disagree with uncomfortable facts are more likely to acknowledge those unwelcome facts.