Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Protect student privacy: ban eproctoring

baneproctoring.com

Post reply on HN
Print view View original

41–50 of 324 posts

Re: Protect student privacy: ban eproctoring

#41

Post by loktarogar »

undfg wrote: post #29
p49k wrote: post #27

Earlier quoted context omitted.

Your stance is that it’s unaffordable for a university to spend a small amount of money on randomized browser tests, therefore they should spend a large amount of money on intrusive spying software?

I don't believe a public institution would be able to come up with a system like this that works reasonably and have it developed in a reasonable amount of time within a reasonable budget.

In 2009 I developed a system like this as part of my course at the equivalent of a community college. The premise of the class was to develop a real world application employing the skills we had learned so far (project management, programming, etc). The whole class (10 people) participated in building this software.

We spent 6 months on it and as far as I heard was still in use as late as 2015.

You don't have to create a whole system; just a database that prepares variants of questions, and a method to put it into the existing exam software.

Re: Protect student privacy: ban eproctoring

#42

Post by donkarma »

jarcane wrote: post #28
dna_polymerase wrote: post #13

> A federally-funded study found that even the best facial recognition algorithms fail to work on Black and brown people, trans and non-binary people, as well as children and women in general. I assume it is black and brown trans and non-binary people. I don't see an algorithm fail because someone identifies as anything. However, why even mention it here, in this context? It makes the people behind this sound ridicul…

Recognition algorithms often make assumptions about the way people "should" be shaped according to their recorded gender or the way the machine attempts to "guess" said gender, and a clash in expectations gets flagged and generates false negatives because things aren't in the "right" places it thinks they should be. For a really simple example if this problem, the TSA body scanner machines flag on trans people consta…

So we put sex on passports? What is the issue?

Re: Protect student privacy: ban eproctoring

#43

Post by wizzwizz4 »

bitL wrote: post #36

Algo for teachers affected by cheating: 1) make questions brutal and unique (compared to previous year or current groups) if possible 2) make everything open-book 3) set a tough time limit 4) curve the results to a desired grade distribution (that is often imposed from above anyway) 5) use software that can the catch most obvious frauds (i.e. 360 degree room scanning, ban use of phones/second computers etc.) 6) if th…

> i.e. 360 degree room scanning

Tape a smartphone to the laptop screen. Out of view of the webcam at all times. (This isn't hard.) Such room scanning is harmful, and doesn't prevent cheating.

Re: Protect student privacy: ban eproctoring

#44

Post by ShroudedNight »

bitL wrote: post #36

Algo for teachers affected by cheating: 1) make questions brutal and unique (compared to previous year or current groups) if possible 2) make everything open-book 3) set a tough time limit 4) curve the results to a desired grade distribution (that is often imposed from above anyway) 5) use software that can the catch most obvious frauds (i.e. 360 degree room scanning, ban use of phones/second computers etc.) 6) if th…

Do you have personal experience finding success with #6? My limited experience with detecting code plagiarism has been that detecting re-use that has been non-trivially obfuscated is a rather difficult problem.

Re: Protect student privacy: ban eproctoring

#45

Post by wizzwizz4 »

donkarma wrote: post #42
jarcane wrote: post #28

Earlier quoted context omitted.

Recognition algorithms often make assumptions about the way people "should" be shaped according to their recorded gender or the way the machine attempts to "guess" said gender, and a clash in expectations gets flagged and generates false negatives because things aren't in the "right" places it thinks they should be. For a really simple example if this problem, the TSA body scanner machines flag on trans people consta…

So we put sex on passports? What is the issue?

> Often they get it coming and going: a trans femme still marked M will get flagged for having breasts, but if they've changed it to F, they get flagged for their genitals instead.

An F sex marker isn't going to stop this happening.

Re: Protect student privacy: ban eproctoring

#46

Post by mytailorisrich »

Negitivefrags wrote: post #6

I can see why universities want to use a system like this. A friend of mine is a University lecturer for physics and one of his students gave him an invite to the discord channel where half his students are in there sharing all the answers during the final exam. He basically told me there is nothing he can do about it. He talked to the head of department and they basically said if you have proof of specific people ch…

I think the only alternative to eproctoring is standard in-person exams in a room at the school/university.

We can debate how far eprotoring solutions should reasonably go but really if someone is allowed to take an exam remotely alone there has to be a form a surveillance.

This can include providing an approved, locked laptop with monitoring sofware instead of letting students use their own devices with eproctoring software but that obviously comes with an additional cost.

Re: Protect student privacy: ban eproctoring

#47

Post by wizzwizz4 »

spoonjim wrote: post #17
dna_polymerase wrote: post #13

> A federally-funded study found that even the best facial recognition algorithms fail to work on Black and brown people, trans and non-binary people, as well as children and women in general. I assume it is black and brown trans and non-binary people. I don't see an algorithm fail because someone identifies as anything. However, why even mention it here, in this context? It makes the people behind this sound ridicul…

Everyone wants a tie-in to the hot button issue of the day. You don’t need to get black trans kids involved to point out all of the flaws in Eproctoring.

To point out enough flaws to disqualify e-proctoring, no. To point out all of them, yes.

Re: Protect student privacy: ban eproctoring

#48

Post by christophilus »

Negitivefrags wrote: post #6

I can see why universities want to use a system like this. A friend of mine is a University lecturer for physics and one of his students gave him an invite to the discord channel where half his students are in there sharing all the answers during the final exam. He basically told me there is nothing he can do about it. He talked to the head of department and they basically said if you have proof of specific people ch…

Spitballing here, but maybe exams need to be reconsidered. What if instead, the professor has a 1:1 conversation with each student, Socrates-style? You’d know pretty quickly whether or not the student was familiar with the material, and if not, where they were weak and needed more focus.

Maybe systematizing everything about education is the wrong approach.

Re: Protect student privacy: ban eproctoring

#49

Post by MattGaiser »

bruce343434 wrote: post #24
spoonjim wrote: post #19

I agree that these e-proctoring measures seem draconian, but the flip side of this is absolutely RAMPANT cheating at universities. I don’t know what kind of fair system lies between these two issues. Also, the domain is poorly chosen, “Bane Proctoring” sounds like a particularly terrifying proctoring service.

> but the flip side of this is absolutely RAMPANT I REALLY doubt this. There's plenty of alternative or hardened testing methods.

Ones that scale?

Re: Protect student privacy: ban eproctoring

#50

Post by anonymousDan »

ShroudedNight wrote: post #44
bitL wrote: post #36

Algo for teachers affected by cheating: 1) make questions brutal and unique (compared to previous year or current groups) if possible 2) make everything open-book 3) set a tough time limit 4) curve the results to a desired grade distribution (that is often imposed from above anyway) 5) use software that can the catch most obvious frauds (i.e. 360 degree room scanning, ban use of phones/second computers etc.) 6) if th…

Do you have personal experience finding success with #6? My limited experience with detecting code plagiarism has been that detecting re-use that has been non-trivially obfuscated is a rather difficult problem.

For me what tends to happen is it becomes easy to catch the weaker students students who are cheating, since they often make weird/unusual mistakes that would be unlikely to occur in the same way (e.g. weird misspellings of certain keywords/variables).
Post reply on HN