Editorial standards
- Last updated October 18, 2026
Summary
On this page
Purpose and limits of the reference
Psych Dictionary is educational reference material about psychology and psychotherapy. It defines terms, describes therapies and techniques, traces their history and reports the state of the research. It does not diagnose, it does not recommend a treatment for any person, and it is not a substitute for care from a qualified clinician. A description of how a technique is used in therapy is a description, not an instruction.
The site collects no health information. An account stores an email address and a list of saved entries, and nothing else that a reader writes.
Structure of an entry
Each entry opens with a one- or two-sentence definition that stands on its own. A panel of facts follows: for a therapy, its tradition, originators, origin and characteristic methods; for a person, dates and main contributions. Therapies, methods and concepts then carry an evidence label and a short evidence summary. The body is organised around what the subject needs rather than a fixed template, and every entry ends with its numbered sources.
Sources and citations
Every empirical or clinical claim carries a numbered citation. Frequency words such as “most” or “often”, effect sizes, prevalence figures and claims about what causes improvement are all treated as claims that need a source.
Sources are chosen in a fixed order of preference: the originating paper or book for history and theory; meta-analyses and systematic reviews for evidence; national clinical guidelines for current practice recommendations, attributed to the guideline. Popular websites, encyclopaedias and commercial blogs are not cited.
Bibliographic details are not typed from memory. Journal articles are cited by DOI, and their titles, authors, journals and page ranges are taken from the Crossref registry. Books are checked against the Open Library catalog by ISBN, or by title, author and year for books published before ISBNs. Web pages are fetched before they are cited. A source that cannot be verified in this way is not cited. Where a registry record is incomplete, the correction and where it was checked are recorded.
Before a claim is attributed to a journal article, the article’s abstract, or the passage concerned, is read. Quotations are used rarely and only from text that has been read in the original, with a page locator.
Books are treated more cautiously. Books are cited for history, theory and how a therapy is described by its originators, never for effect sizes. Some launch entries cite books for descriptive points without a page-by-page check of the book; those citations are recorded in the editorial review notes and are being checked against the books themselves.
Evidence labels
Therapies, methods and concepts carry one of five labels. A label summarizes the research literature for orientation. It is not a recommendation, and it applies to the subject as a whole; where the evidence differs by condition or population, the evidence summary says so.
- Strong evidence
Multiple randomised trials or meta-analyses support the subject for specific, named applications, which the summary names.
- Moderate evidence
Controlled studies support it for some applications, with stated limits.
- Mixed or contested
Findings conflict, effects are small or disputed, or replication has failed.
- Limited or early evidence
Few controlled studies exist; claims rest mainly on theory, case reports or pilot trials.
- Historical theory
Influential in the history of the field, but not supported as originally stated, or superseded.
Effect sizes are reported with their comparison condition, because a therapy compared with a waiting list looks stronger than the same therapy compared with usual care, a placebo or another therapy.
Historical theory, evidence and practice
Entries keep four kinds of statement distinct. Historical theory is attributed to the person who proposed it (“Freud proposed…”). Current evidence is attributed to the research that supports it (“meta-analyses of randomised trials find…”). Contested claims are identified as contested, with both sides cited. Clinical application describes how a therapy is delivered, attributed to its manuals or to practice guidelines.
Authorship and review
Launch entries were written from the sources cited and edited against these standards. Source metadata was verified mechanically as described above, abstracts were read before claims were attributed to articles, and each entry has a review note listing what every citation supports.
No entry has yet been reviewed by a licensed clinician or a subject specialist. An entry shows a reviewer only after a named, qualified person has reviewed it; no review is implied otherwise. The editorial status of each entry is recorded, and entries flagged for update say so.
Corrections
Errors are corrected in place. When a correction changes the meaning of an entry, the “last substantive update” date changes and a short change note records what was corrected. Typographical fixes do not change the date.
Language
Entries use person-first, non-stigmatising language (“a person with schizophrenia”, “died by suicide”). Suicide and self-harm are described factually and without method detail. The register is impersonal: entries describe their subject and do not address the reader.
Independence
The reference carries no advertising and no sponsored content. No therapy, product or provider pays for inclusion or for a label.