What Difference Does It Make?

A few days ago, Retraction Watch published the top ten most-cited retracted papers. I saw this post with a bar chart to visualise these citations. It didn’t quite capture what the effect (if any) a retraction has on citations. I thought I’d quickly plot this out for the number one article on the list.

Retract

The plot is pretty depressing. The retraction has no effect on citations. Note that the retraction notice has racked up 125 citations, which could mean that at least some of the ~1000 citations to the original article that came after the retraction, acknowledge the fact that the article has been pulled.

The post title is taken from “What Difference Does it Make?” by The Smiths from ‘The Smiths’ and ‘Hatful of Hollow’

Advertisements

2 responses

  1. Indeed, this seems much worse than what I would have expected. What can we do to mitigate this problem? At the very least, journals should label retracted references as retracted. That is easy to automate, and I see no reason against it. It may also provide added incentives for authors to check their references so they do not substantiate their claims with too many retracted papers.

    1. I ike that idea and it should be easy to implement. Dave Bridges suggested via Twitter that the journal could easily scan the reference list for retracted papers. I think the context is important to a citation. It’s OK to cite a retracted paper if the reason is discussing why it’s flawed or fraudulent. Obviously, though if the citation is the basis for the current study then this is totally inappropriate.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: