A nurse holding a Pfizer-BioNTech COVID-19 vaccine. (Credit: Joe Raedle/Getty Images)
For many years now, popular trust in science has been in decline. That conclusion may not pass a rigid science review since most of the evidence comes from polling, which — according to Google’s AI review — lacks scientific rigidity. When asked whether polling is a science, Google AI responded with what looks like a consensus view: “It is both a science and an art.” The science involves using “statistical methods, sampling techniques and social science principles to design and conduct polls that accurately reflect public opinion.” Then comes a big cloud of doubt: “However, the interpretation and application of polling data, as well as the specific strategies used by pollsters, can involve a degree of artistry and political judgment.” Some say polling is “more an art than a science.”
The following is not intended as a put-down of polling, which has its valid processes and uses. The summary definition of polling, however, does somewhat correspond to FP Comment’s standard definition of junk science. Junk science occurs when scientific facts are distorted, risk is exaggerated (or underplayed) and “the science” adapted and warped by politics and ideology to serve another agenda. That definition encompasses a wide range of activities among scientists, NGOs, politicians, journalists, media outlets, cranks and quacks who manipulate science for political, environmental, economic and social purposes. We can now add Artificial Intelligence to the list.
The large proportion of science that flows through to the populations of Canada and all countries can, unfortunately, fall into our definition. The array of ideological forces using science to generate public support for social and political causes — or to defeat the same causes — is sowing increasing confusion and distrust.
In recent years, poll after poll after poll has produced evidence that public confidence in science has been declining. Back in 2020, University of Alberta professor Tim Caulfield noted that fomenting distrust “has become the go-to strategy for selling health products, generating clicks and getting elected.” The doubts grew as the COVID-19 pandemic raged and as Donald Trump and others on both sides of the political fence engaged in pitched battles over vaccines.
The pandemic is said to be a major factor behind the decline of trust in science, especially in the United States where vaccine polarization accelerated with Trump’s appointment of Robert Kennedy Jr. as America’s Secretary of Health, which fuelled more doubt (or so the polls showed), especially on the religious right.
Opinions may be shifting, however. Kennedy’s “Make America Healthy Again” report released last month has been roundly trashed across the political spectrum for its lack of science and false scientific statements. The Genetic Literacy Project called the report a “full-scale assault on science.”
In a new paper — The Strange New Politics of Science — two researchers at the American Enterprise Institute argue that while 76 per cent of Americans still trust science, the number is 11 points below pre-pandemic levels. The authors rightly argue that “the stark polarization of American politics around trust in science not only threatens the legitimacy of particular expert institutions, but also has potentially destabilizing consequences for society as a whole.”
A 2024 University of Waterloo survey report, Trust in Canada, suggested Canadians still hold science in high regard. Despite the pandemic episodes, “scientists (along with doctors and researchers) remain one of the most trusted groups in Canada.” In one pre-pandemic poll, 90 per cent of respondents said they “trusted” and “trusted very much” science-related sources. They were followed by science-based personalities (76 per cent), journalists (56 per cent), government (46 per cent), comedians (31 per cent), religious leaders (25 per cent), bloggers and influencers (19 per cent), and celebrities (10 per cent). At least journalists ranked higher than comedians.
A 2023 Confidence in Leaders survey from Environics found that even during the pandemic, 75 per cent of surveyed citizens still had “a lot or some confidence” in science, far behind NGOs (52), journalists (50), business leaders (42) and politicians (33). Such polling results highlight the indisputable fact that trust in science may have weakened to some degree in recent years, but the cause may be more a function of other messengers and institutions rather than scientists.
Which takes us to the heart of junk science. The junk is not necessarily in the science, but in the various ideological streams through which the science flows. Make no mistake, scientists can have political and ideological agendas, but in open debate the junk can be filtered out.
Through this week’s 27th annual Junk Science Week, various science issues are explored, beginning with Peter Shawn Taylor’s exploration of the questionable science behind the annual bee apocalypse. While a Google AI search question (Are bee populations declining?) will produce various versions of yes, the actual answer is no.
Which is not surprising. A recent headline on a science blog said: “Flood of ‘junk’: How AI is changing scientific publishing.” Another claimed that “AI-fabricated ‘junk science’ floods Google scholar.”
And then there is this story from Nature magazine that merged AI with the artful science of polling: “Is it OK for AI to write science papers? Nature survey shows researchers are split. The poll of 5,000 researchers finds contrasting views on when it’s acceptable to involve AI and what needs to be disclosed.”
Contrasting views in science! Situation normal.
• Email: tcorcoran@postmedia.com
Bookmark our website and support our journalism: Don’t miss the business news you need to know — add financialpost.com to your bookmarks and sign up for our newsletters here.
