What was checked
Function-position wording and relationship recommendations were corrected. Career and stress advice, unsourced public-figure assignments and an INFJ growth entry were also corrected. Practical copy covers English, Simplified/Traditional Chinese, Japanese and Korean, with explicit English fallback elsewhere.
The 16 profile descriptions explain the four preference letters, with concrete experience and counterexample questions. Type, result and comparison pages use the same copy in all 44 interface languages, including matching titles and search summaries. The added translations are AI-assisted and have not received independent native-language or expert review. This is not psychometric validation.
Type-page reflection prompts and the /growth guides now use shared optional practices and observation questions. The five translated guides collect three exercises and two fictional scenarios each. They do not assign function training, stress mechanisms or personal deficits by type; scenarios are labelled fictional. This is an editorial correction, not evidence that the practices are effective.
Eight cognitive-function cards now distinguish sourced definitions from hypothetical examples and optional experience/counterexample questions in five languages, with labelled English fallback. Function nicknames and ability/deficit lists were removed. Type/result, learning-card and comparison displays use explicit theory boundaries; function order and scoring are unchanged.
The learning page now uses 16 explicitly fictional scenarios with alternative explanations and a 12-question terminology/boundary bank. Each eight-question round reports correct-answer counts only. Explanations distinguish source terminology from OpenJung editorial boundaries. Overview, metadata and five-language UI use the same scope, with labelled English fallback elsewhere. This is an AI-assisted editorial revision, not expert review or validation.
The function-stack exercise now identifies matches and unlisted orders only within the site’s unchanged 16-entry mapping. Position names, opposite pairs and disputed tertiary orientation have scoped sources. Five-language instructions and keyboard interactions were checked; this is not a personal assessment or validation of the mapping.
What remains unreviewed
Type nicknames, temperament group names, and other untranslated content have not received a full source-by-source review. Lifestyle lists were removed rather than rewritten; Guess-the-Type scenes remain fictional mapping drills, not function identification. The independent mobile dataset has not been updated. These edits do not establish independent expert review or a validation study.