Culture of Integrity – 2

Other research results we learned about in that online cheating MOOC mentioned in the first part of this discussion include:

• Cheating levels tend to go down as the risk and level of punishment associated with getting caught go up
• If cheating requires more work than simply doing the assignment honestly, people will not go down the cheating path
• Levels of cheating in online courses seem to be no different than traditional classes

The first two lessons listed above seem pretty intuitive, and fit within the framework of cheating being more the result of laziness (coupled with risk adversity) than criminality.

The parity between levels of cheating in the classroom vs. online surprised me, although given how much cheating takes one of the many forms plagiarism (which tends to take place in private), that result makes sense (although it should be noted that general perceptions – especially among those not involved with online education – is that cheating in online courses is higher).

One approach to dealing with cheating (regardless of where it occurs) is to treat it as a policing issue.  This involved giving exams in proctored settings with limits placed on what students can bring into a testing room and careful monitoring of their activities while testing is occurring.  And asking students to submit their essays via a software-based plagiarism detection engine acts both as a deterrent and a means of catching people lifting material from sources and using it without attribution.

Combined with mandatory honor codes and hammer-down punishment to anyone who violates them, these types of activities can drive down cheating rates, but only at the expense of turning the professor into a prosecutor who assumes all of his or her students are guilty until proven innocent.

Such policing strategies are controversial, with many educators preferring more cooperative solutions.  These include spending time at the start of a class teaching the rules of non-plagaristic writing (which can dispel misperceptions that lead to unintentional plagiarism) or asking a class create their own integrity code to which they all agree to adhere.

Finally, challenging and creative assignments (such as joint research projects or assignments with deliverables that cannot be built via cut and paste) make positive use of the laziness factor since “cheating” on such assignments would require more work than just doing them properly.

But both coercive or cooperative anti-cheating techniques assume there exists an authority (the professor supported by TAs and backed up by his or her institution) who is in a position to monitor what is going on in class and what is being submitted for grading.  But in the case of MOOCs, no such hands-on, omnipresent authority exists.

Which is why MOOC testing tends to be auto-scored, and delivered through systems that place no controls over who is taking a test and what they can have in front of them while taking it.  And the scale of MOOCs makes peer grading the only practical means for scoring assignments (such as essays) that rise above the level of a multiple-choice question.

Under such conditions, anti-cheating technology such as plagiarism detection systems aren’t very practical (unless someone wants to buy licenses for tens of thousands of potential peer graders, and then teach them how to use it).

As for testing, the solution to this problem for many of my MOOC classes has been to make assessments so easy that cheating (vs. taking tests honestly) becomes the waste of time.

In thinking through what integrity measures could work within a MOOC environment, my thoughts turned to a different technology-driven honesty issue: music piracy, specifically how Apple’s iTunes successfully created a legal alternative to piracy sites such as Napster by:

• Pricing music at an appropriate level
• Making iTunes easier to use than piracy-based alternatives
• Leveraging the influence of the company’s legendary hipster-nerd founder Steve Jobs to create a social norm that said pirating a song (vs. just paying a buck for it) was bad karma

If MOOCs were to follow a similar model, options to consider include:

• Creating distinct levels within a course (similar to what my Einstein professor did when he allowed students to join as auditors or as “qualitative” vs. “quantitative” students who would each have different assignments based on different levels of difficulty).   If each level could be associated with its own unique certificate, this might push people into only signing up for the level of work they feel they can handle.
• Making assignments associated with each level fun and challenging.  With a few exceptions, the assessments I’ve been asked to take within my many MOOCs have felt like afterthoughts, which made them easier to not take seriously.  But this also opens up the door to not taking the class seriously.  So a focus on creating unique, interesting and challenging assignments (ideally ones that leverage the crowds taking such huge classes) could help propel MOOCs into their next stage of evolution.
• Finally, a strategy based on “creating karma” needs to focus not on cheating but on the entire ethos of self-propelled learning that undergirds MOOCs and all other forms of free learning.  High-profile, crowd-sourced global initiatives to generate an integrity code for individual classes or for the entire MOOC movement is one way to generate karma.  And rather than an individual professor giving  a lesson on how to write an essay without plagiarism after the problem arises in class, why not get the world’s best teacher on the subject to create a short MOOC that any professor teaching any online course can direct students towards?

As I mentioned yesterday, cheating is really just a manifestation of corner cutting that naturally occurs once someone realizes that the same reward can be achieved by doing the minimum vs. putting as much effort into a course as possible.  So we need to find ways to celebrate those who go beyond the syllabus, who put time and effort into discussion boards, who write peer-graded essays they would be proud to put into a professor’s hands, and who fully understand that cheating on any course (traditional or massive, brick-and-mortar or online) means selling their soul in order to ensure their own ignorance.