One of the most common ways to quantify research output is to study the citations patterns of published articles.
Why?
- Citation = Impact: if an article is cited, that typically indicates that the work represented in that article had some impact on others in the field.
- Simple: there are numerous relatively simple calculations that can be performed using citation data
- Objective: author evaluations are done using easily verified calculations - this greatly diminishes potential biases that may be associated with methods of evaluation that rely on human judgment.
- Convenient: a number of resources exist to provide people with citation data as well as more advanced statistical analyses based on this data.
Why not?
- Reason for Citation: there are numerous reasons why an article may get cited that don't neccesarily indicate impact - e.g., gratuitous self-citation, work portrayed negatively in citing article, etc.
- Language Bias: articles written in a language other than that spoken by an article's author(s) may be ignored in the works cited - this can be a disadvantage if an author works in a field that is dominated by people who speak other languages.
- Degree of Authorship: being listed as an 'author' on an article doesn't indicate the degree of responsibility for the content presented - there can be a great deal of variation in terms of the amount of credit that an 'author' is due.
- Quality of Journal: work of similar quality but published in a more esteemed journal is more likely to be cited.
- Type of Publication: often times review articles are cited more than original research articles even though they typically do not present new findings.
