The patented algorithms Edifix uses to format, correct, and link your references are pretty smart, but they’re not infallible. If your processed references didn’t come out the way you expected, here are some possible reasons why:

Unexpected author list punctuation

✗ The author list has periods in strange places – for instance, between an author’s first name and middle initial (e.g., Nurse, Derek. M.) or between one author name and another (e.g., Nurse DM, Chow C. Poindexter WJ).

✗ The author list is missing initials for some authors (e.g., Birkholtz, A., Oluransi, and B.S. Knight) or has initials only instead of a surname and given name or a surname and initial(s) (e.g., Birkholtz A, Oluransi J, BS K).

Either of these can cause Edifix to fail to process the author list. 

✔ To resolve this problem, identify and fix the author-name weirdness, then resubmit the reference(s) for processing.

Funky titles or title formatting

✗ Single quotation marks / inverted commas are used, instead of double quotation marks or none at all, around a chapter or article title. This can prevent Edifix from correctly identifying the elements of the entry.

✔ To resolve this problem, delete the ‘/’ or replace them with “/” before processing.

✗ An unrecognized journal title (usually a long and complex title, particularly if it contains a comma-separated list) includes a phrase that Edifix misidentifies as the entire journal title (e.g., in the submitted reference “Zimmerman, B. (2018). Hard heads and hockey pucks: A history of concussion protocols in the NHL.” Sports Geographies: An International Journal of Athletic Space, Place and Environment, 13:3, 443–461”, Edifix incorrectly identifies Place and Environment as the journal title). This can cause Edifix to misidentify the journal title (and, by extension, other elements of the reference, notably the article title), which in turn may cause PubMed and/or Crossref checking to fail.

✔ You can’t successfully re-process the problematic entry, but you can prevent future problems with the same journal title by sending us the correct journal title and ISSN at! Our journal title database is continually updated with contributions from our users :)

Unexpected info / info in unexpected place

✗ An entry for a journal article contains a URL that is not a DOI link. This can prevent Edifix from correctly identifying the entry as a journal reference.

✔ To resolve this problem, try re-processing the entry without the URL

✗ Date information is duplicated within an entry (e.g., Duan, L., and C. Farmer. 2016. “Save Percentages in NCAA Hockey: A Meta-analysis.” Hockey Statistics 21 (2016): 4-42.)

✗ Journal volume number and year of publication are identical because the journal uses the calendar year as the volume number (e.g., Bittle ER, Zimmermann JL. (2016) Ideal jam:peanut-butter ratios for hockey excellence: a randomized controlled trial. Sandwich Sci;2016(1):1-15.)

Either of these can prevent Edifix from correctly identifying and reorganizing all the elements of the reference entry. 

✔ To resolve this problem, remove the duplicate date and re-process the reference.

✗ A reference contains atypical page numbering (e.g., 11.4-11.42; 28a-q). This can prevent Edifix from correctly identifying and restructuring the page information, since not knowing for sure what these numbers are causes Edifix to doubt the identity of other numbers in the reference.

✔ There is no immediate fix for this problem, but we do add support for unusual page numbering formats that are reported to us! So the solution here is to contact Support.

Not enough to go on

✗ A reference contains too little information for Edifix to identify what it is and decide what to do with it. 

Edifix is pretty magical, but even Edifix’s magic has limits! If a reference doesn’t have some basic identifiable elements – such as a title and a date – there isn’t much Edifix can do with it. :(

✗ Your Edifix job contains just one reference 

✔ In general, Edifix will do a better job of identifying and restructuring your references when you submit the entire reference list as one job, rather than processing references one at a time.