Comment by neilv | Hacker News Reader

Comment by neilv | Hacker News Reader

The root problem is that a lot of higher education is nurturing a culture of cheaters right now.

Your future doctors, scientists, government officials, etc... will have had to compete and gain coveted academic and career opportunities, in an environment that both has been heavily gamified, and is being overrun by cheaters.

Insulting measures like this TopHat practically endorses the culture of cheating, by telling students that they can't be trusted, and turning into yet another cheating challenge/task.

Schools with any integrity should be bending over backwards to find, nurture, and support students of integrity.

And to save those who only got admitted by being sketchy, but first semester is a chance to unlearn the bad lessons from before.

Not by treating them as criminals to be monitored, but by treating them like the respectable people they should aspire to be, and which the school expects and requires that they be.

And, for any hopelessly shitty students, who fail to honor this first semester extension of trust, the school should smack them to the curb. Lost tuition income, lost named buildings/chairs, and expensive lawsuits from helicopter parents, be damned.

I have an inside perspective on this via an academic integrity company.

A couple weeks ago there was an exam in an R1 institution that double booked the facility so one section did the exam in person on campus and the other did it "from home". The score distribution of the in person exam was a typical bell curve, and the distribution of the online exam looking like a power-law curve with over half the students scoring 100%.

Thankfully this outraged the professor, and through a variety of means (which I will not disclose publicly) over 25% of the students were caught red handed. Actions are being taken against them, though I'm not sure how far they will go. The evidence against them is overwhelmingly conclusive. In some cases the evidence led to more evidence of cheating in other courses. It seems clear that more that 25% cheated, but I guess catching some is better than none.

As someone who is keenly aware of this crisis, I feel tiny bursts of relief when I see these small wins, though it does feel a bit like bailing an ocean with a teacup.

Catching cheaters is easy. The problem is that incentives are not aligned. Students are also customers, and you can’t simply fire a quarter of your customers and stay in business.

I agree with the latter 100% (the biggest problem is incentives), however I don't agree catching cheaters is easy. It's an arms race, and reminds me of catching cheaters in video games.

Using every means available (multiple technologies and in-person proctored exams) cuts down cheating to probably ~2%... But online programs have serious integrity challenges.

It also doesn't help that our outrage-driven media overwhelmingly exposes us to cheaters.

Everyone's heard of Theranos, Enron, Martin Shkreli, and Bernie Madoff. This week, my 70+ year old aunt asked me about Charlie Javice and Frank. Yet, there are thousands of very successful people quietly building their castles who live and die in relative obscurity because their stories just aren't that thrilling.

If you spend a lot of time interacting with people in the latter category, or if you have them as your mentors, then you will be exposed to a model of what success through hard work and integrity looks like. If you don't, then it's very easy to think everyone successful is a cheater, and that cheating is the only way to break the ceiling into success.

It’s not about individual people - it’s just scale, paired with Goodhart's law.

No number in a spreadsheet will tell you who’s the genuine student. The moment you’re ranking like that you lost.

Long term human interaction in reduced groups is far better at creating genuine environments. But of course, that system doesn’t scale, and it’s a breeding ground for nepotism.

In this moral framework, would it be acceptable for the lecturer to take attendance orally, or is that also insulting?

The instructor clearly sets their expectations for attendance (whether it's mandatory, or otherwise), and then just expects everyone to follow that.

It is verification of attendance, specifically, that "endorses the culture of cheating... telling students they can't be trusted, and turning into yet another cheating challenge/task"? If not, what is fair game for verification, in the pursuit of finding students of integrity?

Finding students with integrity is hard now, because the culture is already full of poo.

But one starting point is to communicate that you expect and require integrity, explain what that means, and then expect it. Trying to make metrics or tests or whatever to detect, rate, rank, etc. it just turns it into a game, like the same load of poo.

Though here is one thing you can do. Explain that you expect integrity, and then watch the students raise their hands and ask how they will be tested on this. You say it's expected. Back and forth a few times, until eventually some of them start crying, and then their heads explode, because they can't figure out how to game that. Those students sadly were too far gone.

Then, after that first semester of integrity culture, some of the students who didn't explode will cheat, and they will be expelled with the fury of an angry god, and everyone on campus will know why. News stories will be written, word will spread, college guides will be updated. The next batch of applicants after that will have fewer cheaters than before, and will have disproportionately attracted students who aspire to integrity and who wouldn't have known to apply to this school before the news.

A school with an honor code that students and faculty take seriously wasn't that newsworthy decades ago, but it's news now.

> Explain that you expect integrity, and then watch the students raise their hands and ask how they will be tested on this. You say it's expected. Back and forth a few times, until eventually some of them start crying, and then their heads explode, because they can't figure out how to game that.

This assumes that the students are untrustworthy and the faculty/institution are ultimately trusted. In a world in which that is not true (such as the world that produces the article we're commenting on), and students sometimes encounter problems due to unclear expectations or vague criteria that are not the student's fault, it is not unreasonable for people to ask questions whose goal is to find out the actual non-vague criteria to avoid unpleasant surprises.

By way of one of many examples: many excellent classes encourage students to talk about assignments with each other, as long as the work they turn in is their own. Now consider what happens if a student accustomed to such a policy encounters a class taught with a different policy, where that policy has not been made clear in advance.

Honor codes and integrity are excellent things to enforce. Transparency and crystal-clear criteria are also excellent things to enforce. Not to allow gaming the system, but to ensure the system doesn't game anyone.

> This assumes that the students are untrustworthy and the faculty/institution are ultimately trusted.

True. This proposal requires expecting and requiring the faculty to have integrity.

And you really need the college/university as a whole to commit to this, not just isolated professors, partly so that there can be no confusion by students.

(Some battle-scarred faculty and grad students could tell speak of entire departments that need to be shut down completely, because the administration and faculty are too far gone. I think you could never do this with one of those departments. You'd only get posturing, and the same arrogant and underhanded behaviors as before, and students would briefly be a little confused, but quickly realize that the old sketchy game-playing is still fully on.)

> True. This proposal requires expecting and requiring the faculty to have integrity.

Not just integrity, but also consistency, objectivity, absence of caprice or bias...

Reply on news.ycombinator.com