It's sort of like the issue of postal workers being fired by false software claims: People trust these allegedly empirical systems to accurately judge students, but don't recognize their limits...
It's sort of like the issue of postal workers being fired by false software claims: People trust these allegedly empirical systems to accurately judge students, but don't recognize their limits and even intended functions.
I've seen students complain about Proctorio. For example, there was a student who failed a test for reading her questions out loud to herself. Proctorio is not in the grading business, they're in the evidence business, they flag behaviors for review by the appropriate authority, typically the professor. In many cases, at least any that are relevant to this discussion, the professor goes "Hmm... this got flagged, so they're cheating." I recall a story I read, but unfortunately can't source, where a student once got lucky and convinced Proctorio to release their video so they could review it with their professor, but I'm cloudy on the details.
The issue, I think, is two-fold: The invasion of privacy, EFF's bread and butter, and the misunderstanding of technology: Arguably something EFF, and schools, should work to remedy. This "proctoring dragnet" exists because of a fundamental misunderstanding as to how these systems are intended to operate.
A window losing focus doesn't mean somebody googled the answer, the burden of proof still lies on the proof of existence of a deed, rather than evidence to its lack of existence.
I also can't help but feel extremely lucky that I've managed to avoid using Proctorio in the last year, partly due to my school subreddit maintaining a list of proctorio-free professors. Even then, all but one of the six professors I've had in the past year used our schools Moodle instance, while the school is trying to migrate to Canvas. Both have limited ways of tracking window activity, IIRC.
Blame the tech giants, specifically the ones from the 90's. When educators (especially at the K-12 level) noticed that technology was going to play a dominant role in the future, they asked the...
The invasion of privacy, EFF's bread and butter, and the misunderstanding of technology: Arguably something EFF, and schools, should work to remedy.
Blame the tech giants, specifically the ones from the 90's. When educators (especially at the K-12 level) noticed that technology was going to play a dominant role in the future, they asked the big technology companies for guidance. And they said 'Learn Photoshop, Office, and Oracle.'
That shift happened very quickly around 1996. By 2004, you could see the regression happening. I had switched colleges, and in my Junior year, I was forced to take some computer skills class as a prerequisite for some others, despite having multiple technical courses on my transcript.
They were covering material that I had learned in 4th grade.
I'm worried we've hit a point of no return. The advent of phones and tablets has further abstracted away users from the underpinings of their devices. Schools and workplaces have locked down machines so hard it's difficult to learn things on your own.
There was a hard shift from 'we'll handle this on our own' to 'we'll just buy all our software'. I work for an org that went from using a series of homegrown apps to vendorizing everything. Costs went down for a hot second, and skyrocketed as missing functionality needed replaced. I pitch replacing a vendor solution and get told 'but it's too hard to do it on our own', ignorinn the fact we've basically rewritten half the functionality, both front and back end, because it didn't do what it needed to.
The EFF can't do this on their own. They've been fighting a losing battle against privacy invasion and anti-user practices for decades, and people just don't care.
People take Apple at face value about security and privacy, and ignore the cage around the ecosystem. The majority still don't understand what DRM is and how it is terrible. And in many ways, Apple is the least terrible.
The problem is that we have managers without technical skills. "Every company is a software company" is still true - so if you're in the position to make a decision on outsourcing vs. building...
The problem is that we have managers without technical skills. "Every company is a software company" is still true - so if you're in the position to make a decision on outsourcing vs. building internal apps you aren't equipped to make that decision if you don't have some technical background.
I develop software for a startup. We outsource plenty of tools to 3rd parties. But it's important to know what tools are any good and what tools can't be outsourced. We don't have a devops team and don't want to spend money on devops just yet - so infrastructure is handled by Heroku. More granular decisions come up when deciding what to turn to npm for and what to roll on our own.
I heard the moment something is flagged, the tool prevents taking the test any further by black screening the computer. If that's true, i don't think you're correct.
Proctorio is not in the grading business, they're in the evidence business, they flag behaviors for review by the appropriate authority
I heard the moment something is flagged, the tool prevents taking the test any further by black screening the computer. If that's true, i don't think you're correct.
It entirely depends on the settings the professor used when setting up the test. Proctorio can be as simple as making sure no other programs are running if the teacher wants.
It entirely depends on the settings the professor used when setting up the test. Proctorio can be as simple as making sure no other programs are running if the teacher wants.
Not intending this as an attack towards your statement specifically, but a general observation: I've had a lot of work in the exam spyware space for a while now, and in my experience "depends on...
Not intending this as an attack towards your statement specifically, but a general observation:
I've had a lot of work in the exam spyware space for a while now, and in my experience "depends on the settings" is the go-to response to "isn't this feature bad for privacy/dangerous/basically broken?", from many CEOs in this space.
"Isn't room scanning privacy-invasive?" "That's a setting the university chooses to enable, you should talk to them." As an example, while all too conveniently sweeping under the rug that they wrote the code for the feature and provided the switch to the university in the first place. If they had any concern about student privacy, they have the individual freedom to not implement that feature. In my opinion, "just an option" is only used by these providers to dodge accountability for their broken, invasive software.
Regarding the power to scan students’ rooms via webcam, Mike says: “Well, that’s a setting that the institution turns on. It’s not on by default. We actually don’t have any defaults. They wanted that. If there’s a problem with it, well we’re just the provider, you need to talk to the institution”.
(Worth a read in general, that article has several other comments that I cannot quantify in words; such as "It’s hilarious, students pretending to care where their data goes.")
This CEO is clearly someone who has no idea what they're talking about, or is very obviously speaking out of both sides of his mouth. What a jerk. If there's a feature that can be turned on, then...
We actually don’t have any defaults.
This CEO is clearly someone who has no idea what they're talking about, or is very obviously speaking out of both sides of his mouth. What a jerk. If there's a feature that can be turned on, then the default is off. If it can be turned off, then the default is on. It's pretty clearly got a default one way or the other.
My running theory is that the exam spyware industry self-selects these kinds of horrible people. Mike Olsen in the WSJ, 4/13/21: "The goal of the software isn’t to police or surveil people." Mike...
My running theory is that the exam spyware industry self-selects these kinds of horrible people.
Mike Olsen in the WSJ, 4/13/21: "The goal of the software isn’t to police or surveil people."
suing university faculty for posting links to unlisted YouTube videos, and then, not joking, trying to invoke an equivalent to DMCA 1201 (anti-circumvention), claiming that "unlisting" a YouTube video was some form of DRM that was bypassed (????)
DMCA'ing my classmate's critical tweets that call into question several of Proctorio's claims by just reading the translation files and looking for strings - he is now suing back with the EFF's support.
With all the things I've done and said about them, I'm actually slightly surprised they didn't go after me yet.
That is horrifying. Anyway, this is absolutely a good point. My own experience with this type of software is pretty limited. Just one class that used proctorio for a few tests. I used an old...
It’s hilarious, students pretending to care where their data goes.
That is horrifying.
Anyway, this is absolutely a good point. My own experience with this type of software is pretty limited. Just one class that used proctorio for a few tests. I used an old laptop with a fresh and unused windows install for the class, and wiped it afterwards. After one of the first tests, my teacher sent me an email that can be summed up with: “the software caught you cheating because you looked away from the screen for a bit. Don’t let that happen on the next tests, or I’ll give you a zero”. This is despite the fact that I did a full room scan, and I was simply looking into the sky while thinking, a perfectly normal thing to do during a test. I had an entire email chain with the professor where he weaseled out of his responsibility to actually have reasonable proof that I cheated.
I haven't seen anything to suggest that Proctorio will black out your entire screen for trying to cheat. Proctorio will only kick you out of the test if it is configured to do so, and you violate...
I haven't seen anything to suggest that Proctorio will black out your entire screen for trying to cheat. Proctorio will only kick you out of the test if it is configured to do so, and you violate one of the conditions. Other systems will black out the test if you try do do something "cheaty" like screenshotting, exiting fullscreen, or copy/paste.
The majority of issues are cause by a combination of Proctorio's monitoring and blind faith by instructors who see it as a way to catch cheaters, rather than just monitor for cheating.
None of this is to defend proctorio, but I think it helps to understand the evil you're against.
It's sort of like the issue of postal workers being fired by false software claims: People trust these allegedly empirical systems to accurately judge students, but don't recognize their limits and even intended functions.
I've seen students complain about Proctorio. For example, there was a student who failed a test for reading her questions out loud to herself. Proctorio is not in the grading business, they're in the evidence business, they flag behaviors for review by the appropriate authority, typically the professor. In many cases, at least any that are relevant to this discussion, the professor goes "Hmm... this got flagged, so they're cheating." I recall a story I read, but unfortunately can't source, where a student once got lucky and convinced Proctorio to release their video so they could review it with their professor, but I'm cloudy on the details.
The issue, I think, is two-fold: The invasion of privacy, EFF's bread and butter, and the misunderstanding of technology: Arguably something EFF, and schools, should work to remedy. This "proctoring dragnet" exists because of a fundamental misunderstanding as to how these systems are intended to operate.
A window losing focus doesn't mean somebody googled the answer, the burden of proof still lies on the proof of existence of a deed, rather than evidence to its lack of existence.
I also can't help but feel extremely lucky that I've managed to avoid using Proctorio in the last year, partly due to my school subreddit maintaining a list of proctorio-free professors. Even then, all but one of the six professors I've had in the past year used our schools Moodle instance, while the school is trying to migrate to Canvas. Both have limited ways of tracking window activity, IIRC.
Blame the tech giants, specifically the ones from the 90's. When educators (especially at the K-12 level) noticed that technology was going to play a dominant role in the future, they asked the big technology companies for guidance. And they said 'Learn Photoshop, Office, and Oracle.'
That shift happened very quickly around 1996. By 2004, you could see the regression happening. I had switched colleges, and in my Junior year, I was forced to take some computer skills class as a prerequisite for some others, despite having multiple technical courses on my transcript.
They were covering material that I had learned in 4th grade.
I'm worried we've hit a point of no return. The advent of phones and tablets has further abstracted away users from the underpinings of their devices. Schools and workplaces have locked down machines so hard it's difficult to learn things on your own.
There was a hard shift from 'we'll handle this on our own' to 'we'll just buy all our software'. I work for an org that went from using a series of homegrown apps to vendorizing everything. Costs went down for a hot second, and skyrocketed as missing functionality needed replaced. I pitch replacing a vendor solution and get told 'but it's too hard to do it on our own', ignorinn the fact we've basically rewritten half the functionality, both front and back end, because it didn't do what it needed to.
The EFF can't do this on their own. They've been fighting a losing battle against privacy invasion and anti-user practices for decades, and people just don't care.
People take Apple at face value about security and privacy, and ignore the cage around the ecosystem. The majority still don't understand what DRM is and how it is terrible. And in many ways, Apple is the least terrible.
The problem is that we have managers without technical skills. "Every company is a software company" is still true - so if you're in the position to make a decision on outsourcing vs. building internal apps you aren't equipped to make that decision if you don't have some technical background.
I develop software for a startup. We outsource plenty of tools to 3rd parties. But it's important to know what tools are any good and what tools can't be outsourced. We don't have a devops team and don't want to spend money on devops just yet - so infrastructure is handled by Heroku. More granular decisions come up when deciding what to turn to npm for and what to roll on our own.
I heard the moment something is flagged, the tool prevents taking the test any further by black screening the computer. If that's true, i don't think you're correct.
It entirely depends on the settings the professor used when setting up the test. Proctorio can be as simple as making sure no other programs are running if the teacher wants.
Not intending this as an attack towards your statement specifically, but a general observation:
I've had a lot of work in the exam spyware space for a while now, and in my experience "depends on the settings" is the go-to response to "isn't this feature bad for privacy/dangerous/basically broken?", from many CEOs in this space.
"Isn't room scanning privacy-invasive?" "That's a setting the university chooses to enable, you should talk to them." As an example, while all too conveniently sweeping under the rug that they wrote the code for the feature and provided the switch to the university in the first place. If they had any concern about student privacy, they have the individual freedom to not implement that feature. In my opinion, "just an option" is only used by these providers to dodge accountability for their broken, invasive software.
Source for Proctorio's CEO making a statement like this.
(Worth a read in general, that article has several other comments that I cannot quantify in words; such as "It’s hilarious, students pretending to care where their data goes.")
This CEO is clearly someone who has no idea what they're talking about, or is very obviously speaking out of both sides of his mouth. What a jerk. If there's a feature that can be turned on, then the default is off. If it can be turned off, then the default is on. It's pretty clearly got a default one way or the other.
My running theory is that the exam spyware industry self-selects these kinds of horrible people.
Mike Olsen in the WSJ, 4/13/21: "The goal of the software isn’t to police or surveil people."
Mike Olsen in the Washington Post, 4/1/20: "We're the police."
In addition to him:
With all the things I've done and said about them, I'm actually slightly surprised they didn't go after me yet.
That is horrifying.
Anyway, this is absolutely a good point. My own experience with this type of software is pretty limited. Just one class that used proctorio for a few tests. I used an old laptop with a fresh and unused windows install for the class, and wiped it afterwards. After one of the first tests, my teacher sent me an email that can be summed up with: “the software caught you cheating because you looked away from the screen for a bit. Don’t let that happen on the next tests, or I’ll give you a zero”. This is despite the fact that I did a full room scan, and I was simply looking into the sky while thinking, a perfectly normal thing to do during a test. I had an entire email chain with the professor where he weaseled out of his responsibility to actually have reasonable proof that I cheated.
I haven't seen anything to suggest that Proctorio will black out your entire screen for trying to cheat. Proctorio will only kick you out of the test if it is configured to do so, and you violate one of the conditions. Other systems will black out the test if you try do do something "cheaty" like screenshotting, exiting fullscreen, or copy/paste.
The majority of issues are cause by a combination of Proctorio's monitoring and blind faith by instructors who see it as a way to catch cheaters, rather than just monitor for cheating.
None of this is to defend proctorio, but I think it helps to understand the evil you're against.