This page documents how research articles on this site are sourced, written, and audited. It exists because the section makes claims about retail trader outcomes and behavioral mechanisms, and a reader who follows the citation chain should be able to verify that those claims sit on primary sources read carefully — not on summary sites, abstracts, or restated talking points.

The rest of this page is the editorial standard every article on the section is held to.

Citation density

Every article cites a minimum of four primary sources. A primary source means one of:

The following do not count toward the citation density floor: Investopedia entries, Forbes columns, broker blog posts, AI-generated content sites, listicle aggregators, the abstracts of papers without reading the full text, secondary citations restated by other articles. Articles may link to these sources for context where it is useful, but the load-bearing claims of the article rest on the primary sources.

Citation verification protocol

Every citation is verified before the article goes to publication. Verification means three things, in order:

  1. The cited paper or report is located at a reachable URL. DOI permalinks are preferred; university repository copies, regulator publication pages, and Google Scholar links are acceptable.
  2. The full text of the paper is read — not just the abstract, not a summary on a third-party site, not the citation as it appears in another paper. The specific passage that supports the claim attributed to the citation is located.
  3. The supporting passage is recorded in the editorial notes so a later auditor can re-check it without repeating the search.

The vocabulary drift script (referenced below) automates a small subset of editorial checks. It does not replace this manual gate.

Outbound link policy

Outbound links to citations point to DOI permalinks, university repositories, regulator URLs, or Google Scholar listings of the paper. The following are not acceptable outbound link targets for citations: Investopedia, Forbes, broker-operated blogs, AI-generated content sites, listicle sites, paywalled summaries.

A regulator or academic source behind a paywall may still be cited — the citation footnote names the source so a reader with access can confirm — but the outbound link should go to the canonical landing page (DOI or publisher), not a paywalled mirror.

Empirical boundary conditions

Every article ends with a section titled "Empirical Boundary Conditions." This section names the limits of the cited research. It states the sample population the study used, the context the research was conducted in, the replication status of the finding where that is known, and any gap between the cited research and the trading-specific claim the article makes.

The purpose of this section is not to undermine the article. The purpose is to demonstrate that the research underneath the article was read carefully and its limits are part of the record. A reader who follows the citation chain back should not encounter a surprise about what the underlying study actually shows. Where the gap between the study's context and the trading claim is large enough that the article's conclusion needs to soften, the conclusion is softened.

Forbidden vocabulary

Articles on this section do not use mechanism-overstatement verbs to describe what the product does — "blocks," "prevents," "cannot override," "stops the trade" — unless the article is describing an actual UI state. The product applies friction to plan deviations and records every override. It is not an execution-layer gate.

Articles do not use wellness or coaching register: "control your emotions," "trade with confidence," "master your mindset," "transform your trading psychology." This vocabulary is the trading-psychology genre's house style. It is not this section's house style.

Articles do not adopt the framing that positions the software as the agent when the trader is the agent. The trader takes the action; the system records it. The full forbidden-construction list is in Parts 8 and 9 of the brand positioning lock. The vocabulary drift script enforces a subset of the list automatically.

Locked vocabulary

The section is built deliberately around six terms used consistently across articles:

Articles aim for repeated, consistent use of these terms over a 12–24 month horizon. The vocabulary drift script reports the density of locked vocabulary per article and flags articles where the density falls below a threshold proportional to article length.

Article topics we do not write

The section does not publish articles in five categories. These prohibitions are not v1-only — they are standing rules documented in Part 10 of the brand positioning lock. The forbidden categories are: emotion-control articles, "best trading psychology books" listicles, day-trading-for-beginners content, trade-journaling how-to guides, and behavioral-bias glossaries.

Each of these topics looks like an easy traffic win and fails the brand positioning for documented reasons. Part 10 names the reasons. The diagnostic for evaluating new topic proposals is also in Part 10.

Author byline

Every article carries a named author. The byline format is stable across articles so the reader builds a recognition of the author who is making the claims. This is an E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) practice and a record-keeping practice — the author's name is attached to the citations they verified.

Update cadence

Published articles are reviewed at six-month intervals against the current state of the cited research. If a paper has been superseded by a stronger replication, or if a replication failure has emerged and changed how the original finding is understood, the article is updated and the last_modified frontmatter field reflects the change. The original publish date is preserved; the update date documents the revision.

The same six-month review checks the article against the current state of the brand vocabulary. If an article published before a vocabulary standard was tightened does not meet the current standard, the article is revised at this review rather than left to drift.

This page is itself subject to that cadence. If the standards on this page evolve, the changes are dated and the prior version is recoverable through the article's git history.