The phrase "no evidence" is a red flag for bad science communication (2021)
The phrase “no evidence” is coming under fire as a confusing and often misleading shortcut in science and policy communication, especially in contexts like COVID guidance, masks, vaccine risks, and transmission. Commenters argue that it blurs crucial distinctions between “hasn’t been studied,” “has weak or mixed data,” and “is strongly disproven,” and that it’s easily weaponized for political or PR purposes. Related debates touch on what counts as evidence in science vs law, how Bayesian reasoning and meta-analyses should shape statements of certainty, and whether experts should instead say explicitly what is known, unknown, and how hard anyone has actually looked.
Meaning and Misuse of “No Evidence”
- Many argue “no evidence” is almost never literally true; usually there is weak, anecdotal, or conflicting evidence.
- Others say it’s a term of art meaning either “no compelling/rigorous evidence yet” or “we have looked and failed to find a significant effect.”
- Problem: the same phrase is used both when something is untested and when it is well-tested and likely false, which confuses the public.
- Some see “no evidence” as a political or PR tool to shut down inconvenient ideas, not a neutral scientific statement.
What Counts as Evidence?
- Disagreement over whether testimony (“Crazy Carl saw aliens”) is evidence:
- Bayesian/epistemic view: any information that changes probabilities is evidence, even if weak.
- Stricter scientific/legal views: evidence should be verifiable, reproducible, or at least corroborated.
- Legal vs scientific standards differ sharply; courts often treat testimony as primary evidence, whereas scientists heavily discount it.
- Some emphasize that evidence quality and context (priors, plausibility) matter more than binary “evidence/no evidence.”
Statistical and Terminology Issues
- Frequent misuse of “statistically significant” and “significant” in media; lay readers hear “important” or “big effect.”
- Correct but easily misinterpreted forms like “no significant evidence of a difference” are often simplified to misleading headlines.
- Meta-analyses and conflicting study results in medicine frustrate some, who expect single definitive studies.
COVID, Masks, and Vaccines
- COVID-era messaging with “no evidence masks work,” “no evidence it’s airborne,” or “no evidence for ivermectin” is heavily debated.
- Some defend these phrases as accurate snapshots of then-available high-quality evidence.
- Others argue authorities downplayed early signals (airborne spread, myocarditis risk) or ignored prior knowledge (surgical mask usage, SARS experience).
- Heated sub-thread on mask efficacy:
- One side cites RCTs and Cochrane-style reviews as showing little or modest effect at population level.
- The other side points to studies and expert commentary suggesting small but meaningful benefits, especially when scaled and properly used.
- Disagreement over data quality, cherry-picking, and misinterpretation of major reviews.
Politics, Trust, and Science Communication
- Many comments stress that senior health officials often speak more as policymakers than as scientists.
- “Evidence-based,” “no evidence,” and similar phrases are seen as rhetorical markers that can signal bias, tribalism, or agenda-driven messaging.
- Some argue that refusing to say “we don’t know yet” erodes trust; others think the public demands oversimplified certainty.
- Broader concern that poor critical thinking, polarized media, and “post-truth” attitudes make any nuanced science communication difficult.