“What I find offensive is not that they plagiarized us, it’s that they did it so badly”

Retraction Watch readers may be familiar with the work of Brian Nosek, a University of Virginia psychologist who has taken a tough stance about many of the problems in his field and coordinates the Reproducibility Project. So it must have seemed quite ironic for Nosek and his co-authors to learn today that one of their papers had been outrageously — and badly — plagiarized.

Here’s the abstract of the work by Nosek, Jesse Graham, and others, which hasn’t been published in a journal yet but is posted at Nosek’s website:

Moral dilemmas pitting concerns about actions against concerns about consequences have been used by philosophers and psychologists to gauge “universal” moral intuitions. Although these dilemmas contain no overt political content, we demonstrate that liberals are more likely than conservatives to be concerned about consequences, whereas conservatives are more likely than liberals to be concerned about actions. This effect is shown in two large, heterogeneous samples and across several different moral dilemmas. In addition, manipulations of dilemma aversiveness and order of presentation suggest that this political difference is due in part to different sensitivities to emotional reactions in moral decision-making: Conservatives are more likely to “go with the gut” and let affective responses guide moral judgments, while liberals are more likely to deliberate about optimal consequences.

Compare that to the abstract of “Political Dogma Stroll’s Non-Political Moral Decision-Making Processes – A Quantitative Analysis of Ideological Decision-Making of Liberals and Conservatives in the Western Europe,” by a group of researchers in Pakistan:

Ethical enigma kernelling concerns about actions against concerns about consequences have been dealt by philosophers and psychologists to measure “universal” moral intuitions. Although these enigmas contain no evident political content, we decipher that liberals are more likely than conservatives to be concerned about consequences, whereas conservatives are more likely than liberals to be concerned about actions. This denouement is exhibited in two large, heterogeneous samples and across several different moral dilemmas. In addition, manipulations of dilemma averseness and order of presentation suggest that this political difference is due in part to different sensitivities to emotional reactions in moral decision-making: Conservatives are very much inclined to “go with the gut” and let affective responses guide moral judgments, while liberals are more likely to deliberate about optimal consequences. In this article, extracting a sample from Western Europe, we report evidence that political differences can be found in moral decisions about issues that have no evident political content. In particular, we find that conservatives are more likely than liberals to attend to the action itself when deciding whether something is right or wrong, whereas liberals are more likely than conservatives to attend to the consequences of the action. Further, we report preliminary evidence that this is partly explained by the kernel of truth from the parodies – conservatives are more likely than liberals to “go with the gut” by using their affective responses to guide moral judgment.

But where the overlap is even more unmistakeable in the Material and Methods sections. Graham et al:

Participants were 656 visitors (48% female, median age 30) to the YourMorals.org website; 473 were from the USA, 70 from the EU, 48 from Canada and 39 from other countries. Political identity was self-reported on a 10-point scale that included a 7-point liberal-conservative continuum plus 3 additional options. There were 356 liberals (three scale points, from slightly to extremely liberal), 80 moderates, and 101 conservatives (three scale points). All analyses retained the seven-point strongly liberal to strongly conservative scaling. The 54 “libertarian,” 27 “other,” and 38 “don’t know/not political” were excluded, leaving a sample of 537.

The Pakistan version:

Participants were 656 visitors (48% female, median age 30) to the YourMorals.org website; 473 were from the UK, 70 from Germany, 48 from France and 39 from other countries of the EU. Political identity was self-reported on a 10-point scale that included a 7-point liberal-conservative continuum plus 3 additional options. There were 356 liberals (three scale points, from slightly to extremely liberal), 80 moderates, and 101 conservatives (three scale points). All analyses retained the seven-point strongly liberal to strongly conservative scaling. The 54 “libertarian”, 27 “other”, and 38 “don’t know/not political” were excluded, leaving a sample of 501.

There are some interesting differences. “France” becomes “Canada,” for example. And 656-54-27-38 = 501 in the new version which is, well, wrong.

Nosek and his collaborators tried emailing the editor listed for the relevant part of the journal, but the email bounced. Graham, the corresponding author, emailed three other associated people, but hasn’t had a response yet.

The journal is published by CSCanada, which runs 15 journals and is on Jeffrey Beall’s list of “potential, possible, or probable predatory scholarly open-access publishers.”

Nosek, to his credit, had a good laugh about the episode.

The fact that they said they created the data from MyMorals.org, that’s our website! It’s comical, it’s so extreme.

We told him it reminded us of being hit with a false DMCA notice by a site that had plagiarized ten of our posts about Anil Potti.

So how did this happen? What seems most likely is that someone took the paper from Nosek’s personal website. The study had been submitted to a journal but not accepted, so the team has been planning new experiments to shore up the paper.

Does it concern him that the kind of openness he champions could be a problem?

My feeling is the opposite. With radical openness, all of this is so easily detected.

In fact, as we often see, it was the fact that the plagiarists cited Nosek and his colleagues’ other work that led to a Google Scholar alert about the new study.

What I find offensive is not that they plagiarized us, it’s that they did it so badly.