A Scientific American article on the prevalence of plagiarism and contract cheating in journal articles.  The tl;dr version lies near the end of the article:

“Now that a number of companies have figured out how to make money off of scientific misconduct, that presumption of honesty is in danger of becoming an anachronism. ‘The whole system of peer review works on the basis of trust,’ Pattinson says. ‘Once that is damaged, it is very difficult for the peer review system to deal with.'” 

Very sad. The only heartening thing about all this is that there are now thousands of scam journals (I think I now get at least half a dozen solicitations from these every day that I have learned to junk immediately) who would be more than willing to publish such articles. I rather like the idea that worse than useless fraudulent articles might get published in worse than useless scam journals. A nice little self-contained economy. Unfortunately, some of the cheats target real journals with real reputations and, worse, may be believed by genuine researchers who are taken in by the lies they purvey, endangering the whole academic research endeavour. Apparently the going price for that in China is around 93,000RMB, or $15,000.

This is very much like the issue we face in course assessment too. In some of my own courses I have designed what I reckon to be virtually foolproof methods of preventing most forms of cheating. They mostly work pretty well, but they don’t cope much better with contract cheating than more traditional assignment/exam based courses. My only partial solution to that problem is to try to price cheats out of the market: most of my courses have to be done from start to finish in order to pass, which is a lot more time consuming than writing a few boilerplate essays, exams or exercises. For assignments and exams on most courses you can get a passing grade for as little as $5, if you are willing to take the risk. The risk of discovery is very high because the essay mills tend to plagiarize or self-plagiarize (well, they are cheats – caveat emptor!) and, due to the semi-public nature of cheating sites, it is just as easy for us to discover students seeking ghost writers as it is for them to seek a ghost writer. In fact, when we find such sites, we tend to pass on our findings to colleagues in other institutions, a nice example of informal crowd-sourcing. However, I am absolutely sure some do get away with it, and it makes little or no difference whether teaching is online or face to face. There’s an example of contract cheating in exams in today’s news, but it is hardly newsworthy, apart from that it is endemic. Beyond contract cheating, I also know that some students have family members or friends who are motivated to ‘help’, sometimes quite considerably. There was a charmingly improbable example of a mother sitting her daughter’s exam a while back, for instance. 

I suspect that the ultimate solution to this in the case of courses is structural, not technological nor even directly pedagogical. We are in an un-winnable arms war in which everyone loses as long as the purpose of courses is seen to be to get accreditation, rather than to enable learning. As long as a grade sits enticingly at the end of it, that will inevitably cause some students to seek shortcuts to getting it. Cheats destroy the credibility not just of their own qualifications but those of every other student who has honestly run the course. If we got rid of grades altogether, cheating during the learning process would dry up to the merest trickle (though, bizarrely, might not go away altogether). Making accreditation a separate issue, completely disassociated from learning and teaching, would allow us to concentrate our firepower on preventing cheating at the point of accreditation rather than distracting us during a course, so we could make our courses far more engaging, enjoyable and useful: we could simply concentrate on pedagogy rather than trying to design cheating out of them. For the (entirely separate) accreditation, we could let rip with all the weaponry at our disposal, of course: biometrics, Faraday cages, style detectors, plagiarism detection tools and all the multifarious technologies and techniques we have developed to attempt to thwart cheats could be employed with relative ease by specialists trained to spot miscreants. Better still, we could use other means of proving authenticity such as social network analysis combined with public facing posts, or employer reports, or authentic portfolios created over long periods with multiple sources of authentication. This would also have the enormous benefit of largely solving what is perhaps the biggest challenge in all of education, that of motivation, getting rid of the extrinsic driver that eats at the soul of learning in our educational systems. It would also allow learners to control how, when, with whom and what they learn, rather than having to take a course that might bore them or confuse them. They could easily take a course elsewhere – even a MOOC – and prove their knowledge separately. It would make it easier for us to design courses that are apt for the learning need, rather than having to fit everything into one uniform size and shape. It would also overcome the insane contradiction of teachers telling students they have failed to learn when, quite clearly, it is the teachers that have failed to teach. Athabasca does, of course, have the mechanisms for this, in its PLAR and challenge processes. It could easily be done.

A similar solution might work, at least a little, for journal cheaters. There are different cultural norms around cheating in China, as I have observed previously, that perhaps play a role in the preponderance of Chinese culprits mentioned in the article, but a lot of the problem might be put down to the over-valuation of publication for career progression, prestige and reward in that country. If the rewards and reputation were less tightly bound to publication and more intrinsic to the process, we might see some improvement. This could be done in many ways: for instance, greater value could be given to internal dissemination of results,  open publication (inherently less liable to fraud thanks to many eyes), team work, blogging, supervisor reports, peer review (of people, not papers) and citations (though that is inevitably going to be the next easy target for fraud, if it is not already, so should not be treated too seriously). There are lots of ways to measure academic value apart from through numbers of publications, many of which relate to hard-to-spoof process rather than an easily forged product. The worrisome trend of journals charging authors for publication is an extremely bad idea that can only exacerbate the problem: publication becomes a commodity that is bought and sold, of value in and of itself (like grades) rather than as a medium to disseminate research.

These are sad times for academia, eaten from the inside and out, but they also present an opportunity for us to rethink the process. The standards and values that have evolved over many centuries and that once stood us in good stead when adult education was an elite affair just don’t apply any more. What our forebears sought in opening up academia was to expand the reach of education to all. Instead, we turned it into a system to deliver accreditation. That system is on a self-destruct course as long as we continue to act as though nothing has really changed. 

Address of the bookmark: http://www.scientificamerican.com/article/for-sale-your-name-here-in-a-prestigious-science-journal/?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%253A+ScientificAmerican-News+%2528Content%253A+News%2529

Very interesting, carefully performed and well articulated study that seems to suggest that showing students their data from early warning systems (learning analytics systems designed to identify at-risk student behaviours, usually through their interactions, or lack of interactions, in a learning management system) generally has a negative impact on their intrinsic motivation.

This is pretty much what one might expect because, as the researchers suggest, it inevitably shifts the focus from mastery to performance, and away from doing something for its own sake. This is probably among the worst things you could do to a learner, so it is not a trivial problem. It doesn’t negate the value of an EWS when used as intended, to help identify at-risk students and to focus tutor attention where it is most needed. I believe that an EWS can be very useful, as long as it is used with care (in every sense) and the results are treated critically. But it does raise a few alarm bells about the need to educate educators not just on the effective use of EWSs but on the nature of motivation in general. 

Address of the bookmark: http://www.sciencedirect.com/science/article/pii/S0747563214003793#

I’d not come across this (2000) article from Vrasidas till now, more’s the pity, because it is one of the clearest papers I have read on the distinction between objectivist (behaviourist/cognitivist)  and constructivist/social-constructivist approaches to teaching. It wasn’t new by any means even 15 years ago, but it provides an excellent overview of the schism (both real and perceived) between objectivism and constructivism and, in many ways, presages a lot of the debate that has gone on since surrounding the strengths, weaknesses and novelty of connectivist approaches. Also contains some good practical hints about how to design learning activities.

Address of the bookmark: http://vrasidas.intercol.edu/continuum.pdf