Sometimes even very good data can fall flat on unreceptive audiences.
When communicating data, it’s important to start with the right expectations. Here are 2 surprising barriers to consider:
- The majority of the population struggles with numeracy and graphicacy. This isn’t just people who claim they’re “bad at math,” it’s a problem for even well-educated professionals (doctors included). In some studies, roughly 6 in 10 people struggled to identify basic trends in a line graph src.
- When we look at data, we see what we want to see. In a 2017 study, Dan Kahan and friends recruited a group of highly-numerate participants and presented cohorts with (hypothetical) data describing the efficacy of a (fictional) skin cream. As expected, the group found easy consensus and interpreted the data “correctly.” They then took the same dataset and gave it a different narrative. Instead of describing the efficacy of skin cream for solving rashes, the experiment cohort was told the data described the efficacy of various gun violence interventions. Their findings: Even for this highly numerate crowd, responses to the data became polarized along particpants’ party lines. That is, the data told them what they already believed. Another study found that the higher an individual’s IQ, the better they are at coming up with reasons to support a position—but only a position that they agree with.