Ranking Methodology

Counted citations are those which cite journal volumes published in the preceding eight years. The reason for this limit is to prevent a bias in favor of long-published journals. Thus the study is concerned only with citations to current scholarship. The search results give only the number of citing documents, and do not show where a citing article or case cites to two or more articles in a cited legal periodical. Sources for the citation counts are limited to documents in Westlaw's JLR database (primarily U.S. articles), and in Westlaw's ALLCASES database (U.S. federal/state cases). The searches conducted in those databases generally look for a Bluebook format in use in the U.S. (volume journal [page] year) but the searches are flexible in allowing the year to occur within 8 words of the journal name.

The list includes periodicals that began publication after the survey period began. Rank results based on total citation counts are unfair to those periodicals, and whenever a journal recently began publication a warning has been supplied next to the periodical name in the form of a parenthetical date such as "(2001- )". Both impact-factor and combined-score rankings do make an allowance for how recently the journal began publishing. Legal periodicals which appear to have ceased publication (even though they were published during a part of the survey period) are excluded.

The "Journals" column(s) shows the number of articles that cite to each journal (within our date period) that were found in the full-text Westlaw journals database "Journals and Law Reviews (JLR)". To see what sources are included in the JLR Database see the Westlaw description at http://goo.gl/2G10s. The scope note in that Westlaw description describes the JLR content as, "The JLR database contains documents from law reviews, CLE course materials, and bar journals. A document is an article, a note, a symposium contribution, or other materials published in one of the available periodicals".

The "Cases" column(s) shows the number of cases that cite to each journal (within our date period) that were found in the full-text Westlaw state and federal case database "Federal & State Case Law (ALLCASES)". To see what sources are included in the ALLCASES Database see the Westlaw description at http://goo.gl/iqnkD. The scope note in that Westlaw description describes the ALLCASES content as, decisions from the "U.S. Supreme Court, courts of appeals, former circuit courts, district courts, bankruptcy courts, former Court of Claims, Court of Federal Claims, Tax Court, related federal and territorial courts, military courts, the state courts of all 50 states and the local courts of the District of Columbia."

Comparisons between older and newer annual surveys cannot be made precisely. Although the number of years covered by each ranking column is an identical 8 years, the size of the JLR and ALLCASES databases for each of the rotated periods increases in size a few percent in each later period. Thus, a small percentage increase in the number of documents citing to a journal may be accounted for by an increase in the total number of documents in the database. The composition of the JLR database is also subject to change with full-text periodicals being added or dropped. For example, should Westlaw add a few new Canadian journals to the JLR database that could have a strong impact on total citation counts to Canadian law journals.

Searches for citing documents (taking the 2002-2009 period as an example) usually look for citations within the full-text articles/cases that have one of the volume numbers published for the journal from 2002 onwards (i.e. where the journal has labeled the issue as 2002..2009), followed immediately by the journal abbreviation/name, and within 8 words a year designation of 2002-2009. A further condition is that any document (case or article) in which such a citation occurs must be dated (in Westlaw's 'date' field) as 2002-2009, and must have been added to the Westlaw JLR or ALLCASES database during the years 2002-2009. These dates should be adjusted for an explanation of other rotated 8 year periods.

A typical looking search in Westlaw would be:
text(65 66 67 68 69 70 71 72 +1 "alb.l.rev." "albany.l.rev." (alb albany +2 "l.rev." "l.r." (l +1 rev review) (law +2 rev review)) /8 2002 2003 2004 2005 2006 2007 2008 2009) & DA(>2001 & <2010) & AD(>2001 & <2010)
Anyone wishing to see the text of the search run for a particular journal in the JLR and ALLCASES databases should send a request to millersc@wlu.edu

As the searches are full-text searches they are naturally prone to some error due to variant citation form in the citing cases/articles. Searches for citation patterns roughly follow the Harvard Blue Book format, usually, VOLUME JOURNAL ... YEAR. Citations that are not in the usual format for legal citations may not have been found. Effort was made to allow for different forms of journal name citations (e.g. allowing for the ALWD citation format) but not all can be retrieved. The citation counts for citations to non-U.S. periodicals are likely to be less accurate than those for U.S. periodicals because non-U.S. legal citation formats are often severely abbreviated.

Westlaw's treatment of periods and spaces within character strings is difficult to understand and is part of the reason why so many alternate cite forms were used.
Quotes are not used around embedded periods, for example, in looking for "HARV. L. REV.", the search HARV +2 L +1 REV would be used.
After a "word" ending with a period, the word is followed by a +2 as in HARV. +2 L.
After letters with embedded periods or single-letters there is no problem with using a +1 as in L +1 REV. However, a problem occurs if the text incorrectly uses a period after a complete word, such as "Akron. L. Rev.", in that case AKRON +1 L +1 REV would not retrieve the cite. Therefore it is a better practice to use a +2 after words longer than one character.