How to tell what is true

The scientific method is not a rigid checklist — it's a disciplined cycle of observation, hypothesis, prediction, experimentation, and revision. Its power comes from its willingness to be wrong. A theory that cannot be disproven is not science; it's dogma.

Hypothesis formation starts with an observation and produces a testable prediction. "This drug reduces symptoms" becomes "patients receiving the drug will show measurably lower symptom scores than a placebo group after 8 weeks." The specificity matters — vague claims resist testing.

Experimental design controls for confounding variables. Randomized controlled trials (RCTs) are the gold standard because they isolate the variable being tested. Double-blinding ensures neither participants nor researchers know who received the treatment, eliminating placebo effects and observer bias.

Peer review and replication are the immune system of science. Before a finding enters the body of accepted knowledge, it must survive review by independent experts and ideally be reproduced by other labs. A single study proves little. Consistent replication across different teams, populations, and methodologies builds confidence.

Consensus emerges when the weight of evidence converges. It's not a vote — it's the product of hundreds or thousands of independent investigations reaching the same conclusion. When 97% of climate scientists agree on anthropogenic warming, that figure represents a mountain of converging evidence, not a poll.

Skepticism is not cynicism. A healthy skeptic doesn't reject everything — they demand proportionate evidence before accepting claims. The goal is calibrated doubt: more skeptical of extraordinary claims, less skeptical of well-established ones.

Scientific skepticism asks: "What's the evidence? Has it been independently verified? Does the claim rely on a plausible mechanism?" This is different from conspiratorial thinking, which often inverts the burden of proof — treating the absence of disconfirming evidence as confirmation, and interpreting any counter-evidence as part of the conspiracy.

Watch for the asymmetry of skepticism. If you apply rigorous scrutiny to claims that challenge your beliefs but accept confirming claims uncritically, you're not being skeptical — you're being biased in the name of skepticism. True skeptics apply the same evidentiary standard regardless of whether the conclusion is comfortable.

A good question to ask yourself: "What evidence would change my mind?" If the answer is "nothing," you're not holding a belief — you're holding a dogma. Every genuine belief should have conditions under which it would be revised.

Epistemology is the study of knowledge — how we know what we know, what counts as justified belief, and how we should update our understanding when new evidence arrives. You don't need a philosophy degree to practice good epistemology. You just need a few core habits.

Start with proportioning belief to evidence. Extraordinary claims require extraordinary evidence — a principle that applies equally to miracle cures and conspiracy theories. If someone tells you a widely-studied vaccine is secretly dangerous, the evidence bar should be extremely high, because decades of safety data say otherwise.

Practice updating beliefs incrementally. Good reasoners don't flip between certainty and doubt — they adjust confidence levels as evidence accumulates. A single study is suggestive. Multiple independent replications are compelling. A meta-analysis of dozens of studies approaches settled knowledge.

Finally, distinguish between belief and justified belief. Everyone holds beliefs. The question is whether those beliefs are supported by evidence, logical reasoning, and expert consensus — or whether they rest on intuition, anecdote, or tribal loyalty.

Social media platforms are the primary information source for billions of people, yet their architecture is optimized for engagement, not accuracy. Understanding how these systems work is essential for navigating them without being manipulated.

Algorithmic amplification is the core mechanism. Platforms surface content that drives engagement — clicks, shares, comments, time-on-page. Emotionally charged content, especially outrage and fear, consistently outperforms nuanced analysis. False claims that trigger strong reactions spread faster than accurate corrections that don't.

Filter bubbles and echo chambers emerge when algorithms learn your preferences and feed you more of the same. Over time, you see an increasingly narrow slice of the information landscape — one that confirms your existing beliefs and hides contradicting evidence. The experience feels like "everyone agrees with me" when in reality, the algorithm has silenced dissent.

Practical strategies: diversify your information sources deliberately. Follow credible outlets across the political spectrum. Be especially suspicious of content that makes you feel strong emotions — that's the engagement algorithm working. Check claims before sharing. If a story seems too perfect for your worldview, that's exactly when you should verify it.

Pseudoscience is any claim, belief, or practice that is presented as scientific but does not adhere to the scientific method. It borrows the language and authority of science without the rigor — no controlled experiments, no peer review, no willingness to be proven wrong. The result is claims that sound scientific but are not.

Why it matters: pseudoscience causes real harm. People delay or refuse proven medical treatments in favor of unproven alternatives. Public policy gets distorted when decision-makers can't distinguish real evidence from manufactured certainty. Resources get diverted from approaches that actually work. And once pseudoscientific beliefs take root, they're remarkably resistant to correction.

How to spot it: pseudoscience has reliable warning signs. Look for claims that rely on testimonials instead of controlled studies, invoke "ancient wisdom" or "natural" as evidence, resist peer review or claim the scientific establishment is suppressing their findings, use vague or unfalsifiable language, and never acknowledge limitations or negative results. Real science invites scrutiny; pseudoscience avoids it.

The antidote to pseudoscience is not cynicism — it's scientific literacy. Understanding how real science works (messy, incremental, self-correcting) makes it much easier to recognize when something is pretending to be science without doing the work.

Logical fallacies are errors in reasoning that undermine the logic of an argument. They're everywhere — in political speeches, op-eds, social media debates, and advertising. Learning to recognize them is one of the most immediately useful critical thinking skills you can develop.

Research papers are the primary source of scientific knowledge, but they're written for specialists. With a few navigation skills, non-academics can extract the essential findings and evaluate their quality without needing a PhD.

Start with the abstract, but don't stop there. The abstract is a summary — it tells you what the researchers found, but not whether the methodology was sound. Treat it as a table of contents, not a verdict.

Check the methods section for sample size, study design, and controls. A study of 15 participants is suggestive at best. A randomized controlled trial of 10,000 is far more persuasive. Look for blinding (did participants know which group they were in?), control groups, and pre-registration (was the hypothesis stated before data collection?).

Understand p-values and confidence intervals. A p-value of 0.05 means there's a 5% probability the result occurred by chance — not a 95% probability the hypothesis is true. Confidence intervals tell you the range of plausible effect sizes. A statistically significant result with a tiny effect size may not be practically meaningful.

Look at who funded it and check for conflicts of interest. Industry-funded studies aren't automatically wrong, but they deserve extra scrutiny — especially when the funder has a financial stake in the outcome. The "Funding" and "Conflicts of Interest" disclosures are usually at the end of the paper.

Preprints vs. peer-reviewed publications: preprints (posted to servers like arXiv or medRxiv) have not yet been peer-reviewed. They're valuable for speed but should be treated with more caution. During COVID-19, many preprints were cited by media as if they were established findings — some were later retracted or significantly revised.

Not all sources are created equal. A peer-reviewed journal, a partisan blog, and a social media post occupy very different positions on the credibility spectrum. Evaluating trustworthiness is a learnable skill with a few reliable heuristics.

Check editorial standards. Credible publications have fact-checking processes, editorial oversight, and published corrections policies. If a source never issues corrections, that's a red flag — it either never makes mistakes (unlikely) or doesn't care about accuracy.

Follow the funding. Who owns and finances the publication? A news outlet owned by a political party, a "research institute" funded by an industry with a stake in the outcome, or a blog monetized by supplement sales all have obvious incentive structures that should inform how you weight their claims.

Distinguish primary from secondary sources. A scientific paper is a primary source. A news article about that paper is a secondary source — it may accurately represent the findings, sensationalize them, or miss crucial caveats. When possible, read the primary source. When you can't, compare how multiple secondary sources cover the same finding.

Red flags for unreliable sources: no named authors, no editorial policy, heavy emotional language, claims of suppressed truth, absence from other credible outlets covering the same story, and a history of publishing debunked claims.

The case for truth in public discourse extends beyond any single issue. Misinformation erodes democratic institutions, undermines public health, delays action on existential threats, and fractures communities. The resources below explore these consequences in depth — and make the affirmative case for evidence-based reasoning.

For an in-depth exploration of why accurate information matters for democracy, science, and public health, see our Why Truth Matters page.

When you want a second opinion on a specific claim, these independent fact-checkers and outlet-credibility trackers are good starting points. They complement a Epistry analysis — use them to sanity-check sources and see how others have evaluated the same claim.