The advice to do your own research has spread far beyond its origins in genuine inquiry. It now functions as both a slogan in online debates and a sincere recommendation from people who distrust institutions. The intent is reasonable: don’t take claims on authority, verify what you can. The execution, in practice, almost always produces beliefs that are less accurate than what the person started with. The reasons are specific, well-studied, and worth understanding even if you reject the slogan’s worst uses.
Search isn’t research
Typing a question into a search engine and reading the top results feels like investigation. It isn’t. Search algorithms surface content optimized for engagement and keyword matching, not accuracy. Someone researching a fringe claim will reliably find websites, videos, and forum posts confirming the claim, because the people producing that content have learned how to dominate those queries. A person researching a mainstream claim mostly finds mainstream sources. In both cases, the output of “research” is largely a function of which keywords you typed and which platforms you used. This is a known limitation in information science — the data is poor quality, the curation is opaque, and the user has no way to assess the sample. Calling it research borrows the credibility of academic work without doing any of the structural work that makes academic research reliable.
Expertise is real and hard to fake
Evaluating a claim about epidemiology, climate, finance, or law actually requires understanding the underlying field. Specialists spend years building the context to recognize when a study is well-designed, when a chart is misleading, when a source is fringe. Non-specialists reading the same materials don’t have that scaffolding and can’t acquire it through afternoon-length engagement. This isn’t elitism; it’s how every technical field works. A non-specialist who spends a few hours reading on a complex topic typically arrives at a worse position than they’d reach by deferring to a reasonable consensus, because the few hours give them confidence without the calibration. The Dunning-Kruger effect describes this dynamic specifically: small amounts of knowledge produce large amounts of confidence.
Motivated reasoning runs the show
The deeper problem is that people doing their own research almost always start with a hypothesis they want to confirm. The questions they enter, the sources they evaluate as credible, the standards they apply to evidence — all are shaped by the conclusion they’re hoping to reach. Studies of online research behavior find that people exposed to balanced information frequently emerge more polarized, not less, because they extract the parts that fit and discount the parts that don’t. Real research practice involves explicitly looking for evidence against your view, weighing sources by methodology rather than by confirmation, and updating your beliefs incrementally. Almost no one does this without training.
Bottom line
Do your own research isn’t useless advice — it’s just much harder than it sounds. If you genuinely want to evaluate a contested claim, talking to actual experts, reading methodologically strong sources, and explicitly seeking disconfirming evidence is the work. Skipping those steps usually produces conviction without accuracy.
Leave a Reply