Yesterday, halfway through writing my next post, I needed a quotation I had used in an earlier post.  I quickly found the quotation, clicked on the link so that I could check and then cite the original source – and, horror, although part of the passage I wanted to use was still there, – the words of the vital sentence were not. They had been replaced, the evidence  I wanted to support my claim was no longer there.

The quotation in question was from the post Flattering flaws. I was commenting on a press release put out by iThenticate.com, promoting their then-recently published study Survey Shows Plagiarism a Regular Problem in Scholarly Research, Publishing, But Action to Prevent Falls Short. I pointed to several questionable statements in the press release, statements which were not always reflected in the actual study.

The paragraph in question reads:

Editors at scholarly publications were the exact opposite, with a majority reporting routinely checking submitting authors’ work for plagiarism. The web site Retraction Watch estimates that the number of retractions in scholarly publications doubled between 2010 and 2011 (iThenticate Press Release, 2012 December 5).

and, amongst other things, I questioned the second statement. There is no evidence in the study to indicate that “the number of retractions in scholarly publications (had) doubled between 2010 and 2011” – and there was nothing on the Retraction Watch website to suggest this either. Where, I asked, had iThenticate found this statement?

I still don’t have an answer to this question. It might not even be a valid question any more, because the statement is no longer there. Instead, what I see now when I go to the press release is

Editors at scholarly publications were the exact opposite, with a majority reporting routinely checking submitting authors’ work for plagiarism. A Nature article reported a 10-fold increase in the rate of article retraction over the past 20 years.

The statement regarding Retraction Watch has disappeared, is not there. There is no proof that my original statement was correct. I could be accused of telling lies, falsifying evidence, sued for defamation, What’s worse, Internet Archive‘s first record of this page is 27 March, 2014 – by which time the change has been made. The paragraph refers to a claim made in Nature, not the claim claimed to have been claimed by Retraction Watch.  If Internet Archive had made any earlier captures, they had been removed.

Fortunately (your honour), I had scrapbooked the original press release; I have a copy on my hard-drive. There, phew, we see

Editors at scholarly publications were the exact opposite, with a majority reporting routinely checking submitting authors’ work for plagiarism. The web site Retraction Watch estimates that the number of retractions in scholarly publications doubled between 2010 and 2011.

just as I claimed in my original post – which was published on 22 March 2013.

For a moment, I was reminded of the scene in George Orwell’s 1984, where, in the middle of a demonstration against Eurasia, the enemy suddenly became Eastasia, and without a blink, the crowds changed the object of their hatred. Eurasia was the ally, and always had been.

Change the record, nobody will know – especially in this digital age when we keep all our records in the cloud, when our secrets can be shared behind our backs, where files are corrupted – not necessarily for corrupt reasons – and our memories are changed and/or disappear. At least, in the scientific community and often elsewhere, when edits are made, when mistakes are corrected, there is a note in the paper that corrections have been made.

Memory hole, NewSpeak, DoubleThink …

I’ll get off my soapbox, to point to one more twist to this tale.  For whatever reason, iThenticate has changed their statement, no longer citing a claim which they attributed to Retraction Watch. Now they state in their press report (originally) published in December 2012)  “A Nature article reported a 10-fold increase in the rate of article retraction over the past 20 years.”

What the article in Nature (published 1 October 2012) says, is

The latest study shows a ten-fold increase (to about 0.01%) in the proportion of papers retracted owing to fraud since 1975.
Zoe Corbyn, Misconduct is the main cause of life-sciences retractions, Nature, 1 October 2012.

What the “latest study” on which the Nature article was based says is

The percentage of scientific articles retracted because of fraud has increased
∼ 10-fold since 1975
Ferric C. Fanga, R. Grant Steenc and Arturo Casadevall, Misconduct accounts for the majority of retracted scientific publications, PNAS 109 (42), 16 October 2012, doi: 10.1073/pnas.1212247109.

One of the questions I raised in my original blog article was to ask whether and how the (alleged) doubling of the number of retractions compared against any increase in the number of articles published? Was there any change in the ratio of number of retractions against number of papers published?

Not only can the same kinds of question be asked of iThenticate’s revised claim regarding the rate of retraction.  We can also ask, does it make a difference to the shock-horror alarm generated by the factoid regarding the “10-fold increase in the rate of article retraction over the past 20 years” if we realise that the ten-fold increase is over 36 years, not over 20 years?

You can prove anything with statistics, especially if you misreport or make them up.  Hey, iThenticate, is another retraction called for?

(Addendum: the article I was writing, before I got diverted … coming soon!)