Citation counts are a quick, if simplistic, indicator of academic impact.
But how many citations are good quality, and how many are low quality, or ‘junk’ citations? I decided to find out.
Using Google Scholar, I reviewed the citations for two of my academic peer-reviewed journal articles. The articles had a combined count of 110 citations.
I couldn’t obtain copies of seven citing publications, even with help from a university research librarian.That left 103 verifiable citations, for which I identified eight distinct types.
- Misquotation. One paper purportedly quoted a passage of text from my article: however, the words quoted did not come from my article, nor from any article I have written.
- Misrepresentation. Twelve publications misrepresented the contents of my article. For example, a number of authors cited my article to support a proposition, when in fact my article did not support their case.
- Bibliography only. Eight publications referenced my article in their bibliography only. So it’s not clear how the author used the article, if at all.
- Ghost citation. In two publications, my article appeared nowhere – neither mentioned in the text nor included in the bibliography – despite Google Scholar recording a citation.
- Duplicate. Fourteen citations were duplicated. Some duplicates resulted from the same publication being loaded to the web with slightly different properties, such as a different title. Some were theses that had also been published as peer-reviewed articles or books. Other duplicates were due to authors replicating almost identical content across different publications.
- Self-citation. Two citations were from publications I had authored or co-authored.
- Relevant in-text reference. 43 publications properly cited my article within their text. Unlike the ‘misrepresentation’ category above, the citation was relevant and representative of the content of my article.
- Correct quotation. 20 publications correctly quoted a passage from my article.
Identifying the junk
How to group these eight types of citations based on quality? I decided to adopt three groupings: junk, quasi-junk, and good quality.
- Misquotation, misrepresentation, and bibliography only (1 to 3 above) I count as ‘junk’.
- Ghost, duplicate, and self-citation (4 to 6 above) I count as ‘quasi-junk’.
- Only the last two citation types, relevant in-text reference and correct quotation, I count as good quality.
Based on these groupings, of the total citations recorded for my articles in Google Scholar, 38 percent were either junk or quasi-junk, and 62 percent – just under two-thirds – were good quality.
I wonder how common that proportion is?
