The Problem of academic Self-Citation
Pete Moore
Academic Training | Researcher Development | Papers | Thesis | Grants | Posters | Output & Impact | Multi-Published Author | Int. Speaker | #ThinkWriteModel
A couple of days ago Nature published an article on academic 'self-citations'. It's not long, and is well worth reading. https://www.nature.com/articles/d41586-019-02479-7
It is a question we frequently face in workshops, and one that needs thinking about carefully.
Lets start by being sensible. If you have been working in an area of research for years and published many papers along the way, then some of your background and discussion will be linked to your previous work. You will have developed methods, and it is important to link to previous papers so that people can spot the continuity. As such, you are introducing self-citation. It's good. It shows that you have a track record and allows people to follow the history of your thoughts and data.
The challenge then is to make sure that the incentives given to citation rates don't encourage bad practice. When, as the article points out, some authors have 50% of their citations coming from their own work eye brows should raise.
Key to the problem, however, it a mechanical use of this data. An author with a surprisingly high self-citation rate could be playing the game or researching an area where few others are working. The author could be failing to keep up with literature or failing to help colleagues know when their work is published so that no one else cites it. It could be so badly written that no one wants to cite it!
The underlying aim of citation metrics is to find simple (maybe simplistic) ways of assigning credit where credit is due. One way of sharpening the tool is to let it identify people who are doing the work that other researchers find helpful in their own lives. To do this, distinguish between the number of citations a paper receives and the number that have come from people not listed on the author line.
Strikes me as a reasonably simple adjustment... but what do you think?
Director of Research, Surrey Sensors Ltd. & Associate Professor at the University of Surrey
5 年All that would happen is that the 'unethical' practice will adjust, so that it continues to deliver maximum citation count under the new constraints. You can't win that way- you'll always create perverse incentives. I'd argue that we need to stop trying to quantify performance in science (and therefore redefine the agenda) with blunt approaches borrowed from business. From Edwards & Roy, Env. Eng. Sci. 34(1), 2017: