17 Evaluating a Website or Publication’s Authority
Authority and reliability are tricky to evaluate. Whether we admit it or not, most of us would like to ascribe authority to sites and authors who support our conclusions and deny authority to publications that disagree with our worldview. To us, this seems natural: the trustworthy publications are the ones saying things that are correct, and we define “correct” as what we believe to be true. A moment’s reflection will show the flaw in this way of thinking.
How do we get beyond our own myopia here? For the Digital Polarization Project for which this text was created, we ended up adopting Wikipedia’s guidelines for determining the reliability of publications. These guidelines were developed to help people with diametrically opposed positions argue in rational ways about the reliability of sources using common criteria.
For Wikipedians, reliable sources are defined by process, aim, and expertise. I think these criteria are worth thinking about as you fact-check.
Above all, a reliable source for facts should have a process in place for encouraging accuracy, verifying facts, and correcting mistakes. Note that reputation and process might be apart from issues of bias: the New York Times is thought by many to have a center-left bias, the Wall Street Journal a center-right bias, and USA Today a centrist bias. Yet fact-checkers of all political stripes are happy to be able to track a fact down to one of these publications since they have reputations for a high degree of accuracy, and issue corrections when they get facts wrong.
The same thing applies to peer-reviewed publications. While there is much debate about the inherent flaws of peer review, peer review does get many eyes on data and results. Their process helps to keep many obviously flawed results out of publication. If a peer-reviewed journal has a large following of experts, that provides even more eyes on the article, and more chances to spot flaws. Since one’s reputation for research is on the line in front of one’s peers, it also provides incentives to be precise in claims and careful in analysis in a way that other forms of communication might not.
According to Wikipedians, researchers and certain classes of professionals have expertise, and their usefulness is defined by that expertise. For example, we would expect a marine biologist to have a more informed opinion about the impact of global warming on marine life than the average person, particularly if they have done research in that area. Professional knowledge matters too: we’d expect a health inspector to have a reasonably good knowledge of health code violations, even if they are not a scholar of the area. And while we often think researchers are more knowledgeable than professionals, this is not always the case. For a range of issues, professionals in a given area might have better insight than researchers, especially where question deal with common practice.
Reporters, on the other hand, often have no domain expertise, but may write for papers that accurately summarize and convey the views of experts, professionals, and event participants. As reporters write in a niche area over many years (e.g. opioid drug policy) they may acquire expertise themselves.
Aim is defined by what the publication, author, or media source is attempting to accomplish. Aims are complex. Respected scientific journals, for example, aim for prestige within the scientific community, but must also have a business model. A site like the New York Times relies on ad revenue but is also dependent on maintaining a reputation for accuracy.
One way to think about aim is to ask what incentives an article or author has to get things right. An opinion column that gets a fact or two wrong won’t cause its author much trouble, whereas an article in a newspaper that gets facts wrong may damage the reputation of the reporter. On the far ends of the spectrum, a single bad or retracted article by a scientist can ruin a career, whereas an advocacy blog site can twist facts daily with no consequences.
Policy think tanks, such as the Cato Institute and the Center for American Progress, are interesting hybrid cases. To maintain their funding, they must continue to promote aims that have a particular bias. At the same time, their prestige (at least for the better known ones) depends on them promoting these aims while maintaining some level of honesty.
In general, you want to choose a publication that has strong incentives to get things right, as shown by both authorial intent and business model, reputational incentives, and history.