This was originally produced in 2015... but kind of horrifies me all the same.
"It's tough because if you tell your kid not to cheat that child is immediately at a disadvantage, and if you tell them to cheat, I think you put them in a longer term disadvantage of greater consequence. I see it as almost a no win situation for parents today, which is very unfortunate."
Some of the comments on the article (from both sides) are informative
Still need time to ponder the meaning.
There's my context. Now my statement(s).
Statement I
The nature of a computer lab has changed significantly since "I was young". The two most meaningful (to me) were the availability of the terminals, and the internet (the instructor controlled both power, and network access). When we attempt to measure individual aptitude, my experience was that half of each exam was completed on paper. Currently, the entire exam is completed on a terminal, and one that is connected to both Facebook and Stackoverflow.
Has "individual competency" changed to the point that I'm measuring it incorrectly?
I think my comments might go more to your points about under-staffing and lack of supervision: it is physically impossible to supervise.
Statement II
I was recently fiddling with an online demonstration of changing of trust dynamics. The basic premise was that trust requires iterative feedback in order to generate knowledge about the counter-party's honesty level. I was also listening to a talk show on which an Economist adamantly stated that cheaters have a competitive advantage.
That computer and cellphone, connected to facebook have significantly reduced the risk associated with cheating. Its more difficult to control, and more difficult to police; this increases the profit margin on cheating. As cheating becomes more profitable, cheaters will become more prevalent, and as it becomes more prevalent, it carries less stigma.
Has morality broken down? Probably not.
Has it's nature changed? Possibly ... maybe even likely
Those rotten kids need to get off my lawn.
Yes, absolutely. Making exams that are short, easy to administer, and easy to grade is a huge part of the problem. The more one automates and labor-saves, and the shorter and simpler the questions, the easier it is to cheat, and the greater the upside for doing so. In the case of comp-sci, I guess cheating (e.g., by looking things up on StackOverflow) is further incentivized by its being exactly how real practitioners would approach solving any knotty problem they hadn't built code for in the past 48 hours.
Maybe the only solution is to take an entirely project-based approach.
My in-class exams are open internet: use stack, use forums, use manuals... The concern is around students paying "tutors" to "help". I don't mind the use of references, I do mind asking someone to write the code for you.
Now we are back to trust. That Facebook window that the weakest student has open... why? When suddenly you pull off a spectacular piece of code, but can't explain it, I have to ask who wrote it.
That is the individual assessment part that I haven't figured out yet. Who am I assessing, you or your tutor?
You want it to be about sharing