Methodology
This project is built on a principle that requires no faith: every claim, every recommendation, and every conclusion must be traceable to a primary source.
Evidence Hierarchy
We use an adapted GRADE methodology. Confidence levels depend not only on the number of studies, but also on their quality, consistency of results, effect size, and real-world applicability.
Meta-analyses and systematic reviews, recognized expert guidelines from WHO, CDC, Cochrane, NHS, and similar organizations.
Multiple high-quality studies with consistent findings, RCTs with adequate sample size.
Several studies with limitations, observational research with controlled variables.
Single studies, correlational data without causal mechanisms.
Contradictory data, speculation, analogies. Such claims are included only with explicit uncertainty labels and limitations.
What We Include in Articles
- Level 4–5 — freely used in main guides
- Level 3 — only with explicit limitations noted
- Level 1–2 — only in uncertainty sections, never as a primary recommendation
Editorial Process
- Define the question — clear, testable formulation
- Gather sources — from whitelist and search via Google Scholar, PubMed, Cochrane Library
- Read primary sources — do not rely on secondary summaries
- Draft an outline — following our fixed template
- Write the draft — with embedded sources
- Verify every fact — matching against the primary source
- Add references and ratings — confidence level for each claim
- Review for clarity — separate facts from opinions
- Publish
- Schedule review — reassess as new evidence emerges
Source Whitelist
Books
- Make It Stick — Brown, Roediger, McDaniel
- Atomic Habits — James Clear (behavior change advice, distinguishing practical tips from scientific evidence)
- Why We Sleep — Matthew Walker (with cross-checking of controversial claims)
- Peak — Anders Ericsson
- Deep Work — Cal Newport
Organizations
- WHO — World Health Organization
- CDC — Centers for Disease Control and Prevention
- NHS — UK National Health Service
- Cochrane Library — systematic reviews and meta-analyses
- APA — American Psychological Association
- American College of Sports Medicine
Research
- Meta-analyses and systematic reviews
- Randomized controlled trials (RCTs)
- Where applicable — cohort studies with confounding control
What We Don't Do
- Do not invent facts
- Do not cite studies we haven't read
- Do not use news articles as data sources
- Do not publish without verifying all factual claims
- Do not publish unchecked machine translations as final content
Periodic Review
Each guide contains a last reviewed date. Guides in fast-moving fields are reviewed at least annually. Guides in established fields are reviewed at least every two years.