Errata are production errors found in a published article or issue.

Errata correct small errors that don't change the overall conclusion or outcome of a paper. This refers to retractions, a procedure issued when errors invalidate the paper's findings. Thus, the entire paper must be retracted so the incorrect findings do not persist.

All errors do not require an erratum to be issued. However, this is a judgment that the Pitt Open Library Publishing staff makes. Here are our guidelines for what errors definitely require errata to be published:

• Changes to the metadata of an article (i.e., wrong author name/initials, incorrect word in title, missing or added keywords)
  • Even if the metadata is correct in the PDF of the article, an errata must be issued to note the changes made to the website metadata that got deposited to Crossref and other indexing services once it was published.
• Something is significantly wrong in the text that impacts interpretation
  • a line on a chart is so thin that it doesn't render in the PDF
  • a reference was included in the text but not in the bibliography
  • a reference in the bibliography was listed under the wrong author
  • a footnote was so long that part of the text spilled off the page
  • the word "there" was misspelled as "their," impacting the interpretation of the sentence

Here are our guidelines for what errors do not require formal errata:

• Errors in the PDF that are actually correct in the metadata, discovered within hours of publication
  • i.e., an author initial was omitted from the PDF but appeared in the metadata
    • This is because the metadata deposit with Crossref is still correct, as the PDF doesn't get deposited. It's only linked. Users will still see the correct PDF.
• Small grammatical typos such as erroneous commas or small typos in the text that don't impact the interpretation of a sentence, discovered within a reasonable time period after publishing
  • i.e., the word "th" appeared and was clearly supposed to be "the"