Editorial standards

This page explains how the educational content across the Shrink Network is written, reviewed, and corrected. It applies to Shrinkopedia and the resource sites. It doesn't describe clinical care at shrinkMD, which follows separate medical and legal standards. Shrinkopedia publishes its own property-level editorial strategy at shrinkopedia.com/editorial-strategy, which extends this network-level standard with the encyclopedia-specific evidence tiers, cluster taxonomy, and Continue Learning block rules.

How content is written

Educational content is written in plain language for the general public. It's calm and specific. It avoids hype, fear, and shame. It explains what's known, and it's honest about what isn't known or still debated.

How content is reviewed

Educational pages are medically reviewed by Shariq Refai, MD, MBA, a board-certified psychiatrist. Each page shows the reviewer's name and the date it was last reviewed, and current reviewers are listed on the medical review board page. A page isn't published with a review stamp until it has actually been reviewed.

What sources are used

Claims are tied to primary and authoritative sources, including the DSM-5-TR, the National Institute of Mental Health, peer-reviewed studies indexed in PubMed, and clinical practice guidelines from organizations such as the American Psychiatric Association and NICE. Sources are listed on the page.

How conflicts of interest are handled

Dr. Refai founded and has a financial interest in shrinkMD, and he earns income from his books and the Unstuck app. Wherever those are mentioned, the connection is disclosed plainly on the same page, in line with FTC guidance. The educational sites take no advertising, no affiliate commissions, and no referral fees. See our disclosures for detail.

How content is corrected

Mental health evidence changes. When it does, pages are updated and the last-reviewed date is changed. Our corrections policy explains how errors are reported and fixed. If you find an error, email [email protected].

Network linking standard

The Shrink Network is an ecosystem of independent sites, each with one job. For that to be more than a claim, the sites have to link the same way. These rules bind every site in the network. They're published here so any editor, contributor, or automated tool can check a page against them.

One canonical definition per concept

The canonical definition for any concept lives on Shrinktionary. Site glossaries elsewhere in the network aren't banned. They just aren't allowed to compete.

Right now, seven sites in the network define the same words: Shrinktionary (where it's the whole job), Shrinkopedia's /terms/, PsychiatryRx's /glossary, AnxietyResource's /glossary, DepressionResource's /glossary, AnxietyResearch's /glossary, shrinkMD's /glossary, and even shariqrefai.com's /glossary. That costs the network three ways. It splits authority, so five thin pages about "rumination" lose to one deep page every time. It creates drift, so the day PsychiatryRx's definition of physical dependence stops matching Shrinktionary's, the network is publishing two answers under one reviewer's name. And it wastes the actual asset: Shrinktionary already publishes canonical concept IDs at shrinktionary.com/concepts.json, so the "one shared graph" story is a claim until the rest of the network uses those IDs.

The rule for every glossary entry on a non-Shrinktionary site:

1. Render the definition, don't rewrite it. Where you need the plain definition of the word, fetch it from concepts.json at build time and render definition directly. Attribute it and link to url. Four sites displaying the same sentence pulled from one source can't drift. Four sites each writing their own sentence and agreeing to stay aligned will drift, and the first anyone hears about it is a reader noticing that two of your sites contradict each other.
2. Add your angle in one or two sentences. AnxietyResource explains what hypervigilance means for someone with anxiety. Shrinktionary explains what the word means. That's the whole shape of the value each site adds.
3. Deep-link home with visible attribution. Use link text like "Full definition on Shrinktionary" pointing at https://shrinktionary.com/terms/<slug>/. The slug is the canonical concept ID from concepts.json. Do not link to Shrinktionary's homepage.
4. Do not target the definition query. Don't title the page "What is rumination?" or "Rumination: definition and meaning." Those belong to the language layer. Title it for your own angle: "Rumination and anxiety," "Rumination in depression," "How SSRIs affect rumination."
5. Add DefinedTerm schema with sameAs. Point at the canonical Shrinktionary URL so search engines and AI assistants see one concept across five pages, not five competing ones. Example:

   {
     "@type": "DefinedTerm",
     "name": "Rumination",
     "termCode": "rumination",
     "sameAs": "https://shrinktionary.com/terms/rumination/",
     "inDefinedTermSet": {
       "@type": "DefinedTermSet",
       "name": "Shrinktionary",
       "url": "https://shrinktionary.com/"
     }
   }

Shrinktionary's reciprocal obligation, already in place: it deep-links into specific pages on your site rather than dumping readers on your homepage. That map is maintained on Shrinktionary and link-checked every quarter. Sibling sites should return the favor.

Concept IDs are the Shrinktionary slugs

Concept IDs are published at shrinktionary.com/concepts.json. Every site uses those IDs when it refers to a concept. When shrinkDaily publishes a lesson on metacognition, the lesson carries a canonical_concept field pointing at the Shrinktionary slug for metacognition. The lesson stays a lesson; the definition stays on Shrinktionary. Two graphs is zero graphs, so we run one.

Cross-links go to pages, not homepages

If a page on one site links to a sibling site about a specific thing, the link goes to that thing's page. A homepage handoff makes the reader start their search over, which is the opposite of a network.

Every site publishes /index.json

Every site in the network publishes an /index.json at its root that lists its canonical pages. Each page item carries its concept_ids where applicable. This is what makes the network machine-readable. ShrinkNetwork.com publishes /index.json as its own contribution and publishes /network.json as the registry of every other site's index.

One Person node

Every site's author, reviewedBy, and creator in schema references {"@id": "https://shariqrefai.com/#person"}. Reviewer identity is single and authoritative across the network.

Roles the network runs on

Every site declares one role so the network can route by function. The current roles: language (Shrinktionary), understand (Shrinkopedia), learn (shrinkDaily), condition (AnxietyResource, DepressionResource), evidence (AnxietyResearch), medication (PsychiatryRx), apply (shrinQ, Unstuck, QuitScrolling), care (shrinkMD), profession (shrinkiatry, shariqrefai.com), and registry (this site).

How claims stay true over time

The network runs a scheduled quarterly re-verification of every legal claim, primary source citation, and outbound link, plus a lighter pre-December pass to catch year-end regulatory changes. Details and cadence are in /network.json. When a page passes verification, its Last reviewed date is bumped. When something fails, the page is flagged and the reviewed date is not bumped until the underlying issue is fixed.