For decades, the scientific community has been nearly unanimous: Climate change is real, it's our doing, and its consequences are likely to be severe. Yet even as it gets more difficult to avoid some of its effects, poll after poll shows that the public hasn't gotten the message. There's very little recognition of how strong the scientific consensus is, and there is a lot of uncertainty about whether it's our doing—and none of the polling numbers seem to shift very quickly.

Over these same decades, there have been plenty of studies looking at why this might be. Many of them have found ways to shift the opinions of study subjects—methods that have undoubtedly been adopted by communications professionals. Yet the poll numbers have remained stubborn. Misinformation campaigns and political polarization have both been blamed, but the evidence for these factors making a difference is far from clear.

A new study offers an additional hint as to why. While polarization and misinformation both play roles in how the public interprets climate science, the biggest problem may be that the public has a very short memory, and anything people learn about climate science tends to be forgotten by a week later.

Time after time

To test people's responses to climate information, researchers gathered a set of materials that had appeared in major publications. Some weren't climate-related and served as a "placebo." Others were coverage of an earlier report from the Intergovernmental Panel on Climate Change. Finally, there was a set of articles that focused on partisan disagreements regarding climate change (as opposed to the scientific content) and a set of opinion pieces that argued against accepting the scientific evidence.

Advertisement

The study focused on creating a bunch of paths through this information, with different readings in consecutive weeks. For example, one group of participants might receive science all the way through, while another might get science one week and then have it contradicted by an opinion piece the week after. The goal was to detect whether exposure to science had a lasting effect or if it could be undercut by either time or misinformation.

The risk here was that having so many potential paths through the information would mean that only a few people went down each particular path, making any results statistically suspect. The researchers overcame this by recruiting a lot of participants—nearly 3,000 people did the entire multi-week process. To do so, they had to rely on Mechanical Turk, a service some users have managed to script. But a number of studies have indicated that Mechanical Turk results have been replicated by in-person studies, so the researchers felt it was sufficiently reliable.

The experiment ran over four weeks. In the first, basic information about the participants' existing beliefs about climate change was established. Afterward, there were two weeks of reading articles, followed by additional polling. Week four simply saw a final poll to determine whether the previous weeks' reading had changed any opinions.