Is it just me or has the US Government been doing a lot of actions recently. Just today I saw two other articles about how the gov is cracking down on Juul, some anti right to repair company, etc. Is it just all getting published now?
Nope. That online test proctoring is the real surveillance, too. This ruling is just going to make edtech companies way less likely to use their product data to support product improvement. It is already super rare in edtech—because you don’t make more money when your product creates better learning! So edtech companies rarely use their existing data.
* Based on my experiences working with many education companies.
What exactly do you mean by "the real surveillance"?
Coerced 'acceptance' into this sort of shit after you sign up for a class is unconscionable. And that's aside the medical disabilities, bathroom needs, being black, being asian... Frankly, being anything not white and no medical issues.
> This ruling is just going to make edtech companies way less likely to use their product data to support product improvement.
The ruling will scare these lousy administrators from even buying seats to the software, let alone using it to "improve educational outcomes".
And to that, I say "Good riddance". Burn that whole surveillance market. Completely.
And yes, I've also worked for a while in an educational setting as a systems engineer. There's a reason I left that space. It's corrupt as fuck, nobody actually cares - its just a bunch of sycophants making non-changes or bad changes and looking at beneath-noise rates of margin of errors.
The FTC signaled this move and others earlier this month.[1]
>“During the pandemic reliance on these technologies really dramatically increased and Americans are now often dependent on these key digital tools and services to navigate their day to day lives,”
>“Children should not be required to sign up for surveillance in order to sign up to do their schoolwork,” Khan said Wednesday.
I haven’t been particularly careful with my online privacy, and I actually click on advertisements on social media occasionally. Typically for clothes or blood bowl related stuff and I sometimes buy the thing I clicked on. So there is no doubt that targeted advertisements work on me, the thing is, however, and this is completely anecdotal, a lot of it is just really bad.
Recently I have been getting a lot of advertisements for guns. I’m Danish and I don’t have a hunting license which means I can’t legally buy guns. This is just one example, others have provoked me so much I’ve actually gone into my advertisement settings to block the sender forever. If that’s the best they can come up with, with all my data, then they can frankly just fuck right off. Especially because it wouldn’t be hard for them to “target” the blood bowl advertisement without tracking my every move considering it’s the only reason I have social media accounts outside of HN and LinkedIn.
What is really interesting to me, however, is that I don’t notice much of a difference in the advertisements on Google and DuckDuckGo, one knows more about my life than I do and the other simply goes “well he searched for blood bowl, maybe I’ll show him some blood bowl advertisements”. In fact I’m only on DuckDuckGo because the Google SEO bullshit has made the front page of Google results useless for almost all of my searches.
I'm more concerned with the huge corporations. I think they are the worst.
Legislation like this actually creates a bar to smaller companies. The big corps are already in, aren't they just passing legislation to prevent competition?
I don't think it's as complicated as that. Simple answer is the scope of the bill is narrow to ensure it passes. Anything bigger and lobbying power becomes an issue.
according to the article it's an FTC policy statement, not a bill. the article says it was voted in 5-0. I could be wrong about this, but are the FTC committee not just stooges of whoever is President? as seen in the whole Ajit Pai Net Neutrality business
What are the ethical implications of NOT using student learning data to support student learning? How will this ruling affect the development of adaptive or personalized learning systems? How will it, for instance, affect the ability to identify learning gaps at scale based on race or gender?
A kid’s world is overloaded with advertising optimization in their games, videos and online experiences. It happens whenever kids don’t announce themselves as kids. I’m worried that this ruling won’t do anything about hyper optimized advertisements but will create meaningful barriers to the data-driven improvement of educational software.
Keep in mind that most edtech companies are decades behind other technology companies in their use of user data. That’s because the kids don’t pay, so optimizing their experiences or learning outcomes don’t meaningfully affect the bottom line. In this context, framing the use of student learning data as “surveillance” is very cynical.
To give an example, this ruling appears to forbid the use of formative assessment data (patterns of what students know or don’t know) to build adaptive learning systems. If so, that would significantly impair edtech innovation.
Seems like you have the exact opposite take here that I do. Rather than allowing surveillance of kids because we are being surveilled by other companies, maybe make that surveillance harder too?
Not to mention that, at least theoretically, you have a choice to interact with the other people surveilling you. Kids have no choice and are forced to go to school and would be under forced surveillance and data harvesting.
I do have a different take. My experience comes from helping edtech companies with data science. So, I get to see many potentially good things that edtech companies don’t do.
I personally think it is ethically imperative to analyze patterns of student performance and motivation within learning software, with the goal of system/software improvement. Why? Because enhancing the efficacy of education at scale is really, really valuable—and I’m not speaking from a finance perspective. In fact, there is very little money to be made by making online education more effective.
I don’t agree that using data for continuous product improvement should be called “forced surveillance and data harvesting.”
I know there is a real issue with advergaming and “user as product.” This is not what I’m referring to. I’m asking for more nuance.
You're free to do whatever research on children's behavior after you get informed opt-in non-coerced consent from the legal representatives of all the participants of the study and approval from a supervising IRB (Institutional Review Board)/REC (Research Ethics Committee), otherwise it should be called "forced surveillance and data harvesting".
This is not a new discussion, the research community has long ago established a near-consensus about what are the appropriate ethical boundaries, permissible tradeoffs, and non-negotiable safeguards when studying human subjects, and this goes waaaaay beyond that.
Please don’t do the extreme outrage machine. Please listen.
Those consensual agreements you speak of specifically include the use of research to support normal educational practices. That’s why conducting research on normal educational practices is one of the few exceptions to the requirements for informed consent in research. That’s because it was understood that there is an ethical trade off —gathering informed consent from kids in education settings is so difficult that they knew requiring consent would prevent the research from occurring. That’s what I’m saying —rather than “black and white” thinking that collecting data from kids is evil, consider the real societal losses from banning research to improve educational practices— including what is now normal online educational practices. This needs nuance.
Yeah with no concerns about the limitations or monetization of the data that the edtech companies are collecting. Data on the habits of children has to be some of the most valuable data in the data-for-sale corporate/black markets.
Not that IRB is too onerous. We still get IRB approval. But approval can come with an exemption from informed consent when it involves research on normal educational practices.
> This is not a new discussion, the research community has long ago established a near-consensus about what are the appropriate ethical boundaries, permissible tradeoffs, and non-negotiable safeguards when studying human subjects, and this goes waaaaay beyond that.
Unfortunately, our definitions are extraordinarily broad and technically almost anything is human research by the letter of the definition. In practice, we don't call everything that interacts with data about people "human experimentation", but figuring out the exact line is very hard.
E.g. I am a teacher. I do things differently from year to year. I do my best to gather the best metrics on this time series data that I can in order to inform my choices on what I consider best practices. Do I need to get an IRB involved? My institution doesn't have an IRB.
I really wish there were high quality controlled experiments in education that I could rely upon. But the vast majority of the educational research is crap.
Think about, say, Khan Academy. I have a hard time imagining a world where interacting with millions of kids on a protocol is OK, but the second we take some basic metrics to try and figure out how well it works, it's bad. I'd arguably ethically one needs to perform a certain basic level of monitoring and optimization of any intervention.
> I know there is a real issue with advergaming and “user as product.” This is not what I’m referring to. I’m asking for more nuance.
Intent doesn't matter unless the data collection includes legally binding provisions to delete all collected data in the case of bankruptcy or acquisition. Otherwise this is just a pile of user data waiting to be scooped up by private equity and sold to third parties.
> In fact, there is very little money to be made by making online education more effective.
Right. So when the ed-tech company gets into financial trouble the data becomes an asset to be sold.
I hear you. Good idea. I think the core idea that you can’t use data to support marketing or sales is right. And, I’d support a blanket ban on selling the data, except when you are selling the company in a manner that continues the educational purposes.
As far as I can tell, this doesn't aim to stop the use of student learning data to support student learning. I think it's about the "other stuff", like selling PII to third parties or using it for targeted advertising.
Then that should clarified by the FTC. As you can see from other comments, it appears that some people want to stop the use of student learning data to support student learning.
Educational systems fundamentally need to collect data on student learning progress. It is how education functions as a system. Lumping together the unethical (selling PII) and the ethical (continuous improvement) is very problematic.
It's clarified in the actual policy statement [0][PDF].
Particularly:
> must not condition participation in any activity on a child disclosing more
information than is reasonably necessary for the child to participate in that activity.
> are strictly limited in how they can use the personal information they collect from children. For example, operators of ed tech that collect personal information pursuant to school authorization may use such information only to provide the requested online education service.
They're not prohibiting the collection of data, they're strictly limiting what can be done with it.
I think this is good middle-ground privacy policy in general.
> must not condition participation in any activity on a child disclosing more information than is reasonably necessary for the child to participate in that activity
That may prevent gathering data about, for instance, their performance on other related skills.
> are strictly limited in how they can use the personal information they collect from children. For example, operators of ed tech that collect personal information pursuant to school authorization may use such information only to provide the requested online education service
How about using their PII to link data sources together? That might be done for the improvement of related systems, rather than to provide the educational service used by the child. I mean it’s arguable, but this could create a real chilling effect.
Why stop at children? College students have been increasingly surveilled during the pandemic and the shift to online education.
The popular tool Lock Down Browser is more or less spyware that takes over your device, monitors every click, and watches every movement. And yet most colleges and universities can’t reliably detect cheating even with this tool.
So what's the magic age at which it becomes morally acceptable to surveil people? Is this a way to ease into the argument that freedom from surveillance is a human right?
I mean to be fair any college exam is physically proctored by TA's and teachers to avoid the same thing before. Though the solutions aren't great there is something to be said for attempting to keep the validity of an exam.
Well designed tests aren't something you can get through with a textbook or a web browser.
I would say the majority of tests are not well designed, and rely on rote memorization and recall. It's disheartening to see schools implement surveillance to compensate.
I would expect a lot of this software locks down your computer in the name of "preventing cheating" but the company's unstated goal is "harvesting more information."
In a similar vein, after 9-11 the airlines loved the idea that they had to start checking the ID of every passenger to ensure it matched the name on the ticket. They said they supported this for "increased security" but the real reason the airlines liked the idea was that it eliminated ticket resale, which had been a major annoyance to the airlines' bottom line for decades.
This is partly the fault of school systems. During the pandemic with online learning they had my kids bring their Chromebooks home to do online classes. Their solution for every issue was to install more and more Chrome extensions and services connected to their Google Account mostly which also synced with their email.
Half of them clearly stated that they were collecting data and yet that school didn't have the kids sign any release from the parents. These are kids under 13 years old BTW. Essentially, the school decided for parents it was okay for random Chinese and other apps to spy on students without the parents permission.
the actual guidelines in the linked pdf are not useful. it does seem to allow, for example, using the content of stuff kids write to train stats models. (writing content = 'reasonably necessary to participate in the service', guideline 1, and training a model isn't 'unrelated to provision of school requested service', guideline 2). and the trained model is no longer PII, but can definitely be used to advertise to kids on youtube. (via awareness of topics they like, for example).
there are no strong anonymity requirements. no discussion of merging student's profile with their google ad profile in N years when they're out of school, or defaulting to being logged in in the browser.
this reads like the ftc has no idea what the harms here are and is unwilling to come down on schools for requesting things that are dystopian, like eye tracking in exams.
51 comments
[ 4.7 ms ] story [ 115 ms ] threadIt is also Supreme Court decision announcing time, they do run on a calendar.
[1] https://www.washingtonpost.com/technology/2022/06/08/ftc-lin...
https://mattstoller.substack.com/
* Based on my experiences working with many education companies.
Coerced 'acceptance' into this sort of shit after you sign up for a class is unconscionable. And that's aside the medical disabilities, bathroom needs, being black, being asian... Frankly, being anything not white and no medical issues.
> This ruling is just going to make edtech companies way less likely to use their product data to support product improvement.
The ruling will scare these lousy administrators from even buying seats to the software, let alone using it to "improve educational outcomes".
And to that, I say "Good riddance". Burn that whole surveillance market. Completely.
And yes, I've also worked for a while in an educational setting as a systems engineer. There's a reason I left that space. It's corrupt as fuck, nobody actually cares - its just a bunch of sycophants making non-changes or bad changes and looking at beneath-noise rates of margin of errors.
>“During the pandemic reliance on these technologies really dramatically increased and Americans are now often dependent on these key digital tools and services to navigate their day to day lives,”
>“Children should not be required to sign up for surveillance in order to sign up to do their schoolwork,” Khan said Wednesday.
[1] https://thehill.com/policy/technology/3516884-ftc-chair-khan...
Recently I have been getting a lot of advertisements for guns. I’m Danish and I don’t have a hunting license which means I can’t legally buy guns. This is just one example, others have provoked me so much I’ve actually gone into my advertisement settings to block the sender forever. If that’s the best they can come up with, with all my data, then they can frankly just fuck right off. Especially because it wouldn’t be hard for them to “target” the blood bowl advertisement without tracking my every move considering it’s the only reason I have social media accounts outside of HN and LinkedIn.
What is really interesting to me, however, is that I don’t notice much of a difference in the advertisements on Google and DuckDuckGo, one knows more about my life than I do and the other simply goes “well he searched for blood bowl, maybe I’ll show him some blood bowl advertisements”. In fact I’m only on DuckDuckGo because the Google SEO bullshit has made the front page of Google results useless for almost all of my searches.
I'm more concerned with the huge corporations. I think they are the worst.
Legislation like this actually creates a bar to smaller companies. The big corps are already in, aren't they just passing legislation to prevent competition?
I thought that was the point the person you responded to was making.
[1] https://fairplayforkids.org/
[1] https://fairplayforkids.org/
A kid’s world is overloaded with advertising optimization in their games, videos and online experiences. It happens whenever kids don’t announce themselves as kids. I’m worried that this ruling won’t do anything about hyper optimized advertisements but will create meaningful barriers to the data-driven improvement of educational software.
Keep in mind that most edtech companies are decades behind other technology companies in their use of user data. That’s because the kids don’t pay, so optimizing their experiences or learning outcomes don’t meaningfully affect the bottom line. In this context, framing the use of student learning data as “surveillance” is very cynical.
To give an example, this ruling appears to forbid the use of formative assessment data (patterns of what students know or don’t know) to build adaptive learning systems. If so, that would significantly impair edtech innovation.
Not to mention that, at least theoretically, you have a choice to interact with the other people surveilling you. Kids have no choice and are forced to go to school and would be under forced surveillance and data harvesting.
I personally think it is ethically imperative to analyze patterns of student performance and motivation within learning software, with the goal of system/software improvement. Why? Because enhancing the efficacy of education at scale is really, really valuable—and I’m not speaking from a finance perspective. In fact, there is very little money to be made by making online education more effective.
I don’t agree that using data for continuous product improvement should be called “forced surveillance and data harvesting.”
I know there is a real issue with advergaming and “user as product.” This is not what I’m referring to. I’m asking for more nuance.
This is not a new discussion, the research community has long ago established a near-consensus about what are the appropriate ethical boundaries, permissible tradeoffs, and non-negotiable safeguards when studying human subjects, and this goes waaaaay beyond that.
Those consensual agreements you speak of specifically include the use of research to support normal educational practices. That’s why conducting research on normal educational practices is one of the few exceptions to the requirements for informed consent in research. That’s because it was understood that there is an ethical trade off —gathering informed consent from kids in education settings is so difficult that they knew requiring consent would prevent the research from occurring. That’s what I’m saying —rather than “black and white” thinking that collecting data from kids is evil, consider the real societal losses from banning research to improve educational practices— including what is now normal online educational practices. This needs nuance.
For more information: https://research.virginia.edu/irb-sbs/normal-educational-pra...
And then you have the audacity to tell us to "don’t do the extreme outrage machine"?
Just wow.
Unfortunately, our definitions are extraordinarily broad and technically almost anything is human research by the letter of the definition. In practice, we don't call everything that interacts with data about people "human experimentation", but figuring out the exact line is very hard.
E.g. I am a teacher. I do things differently from year to year. I do my best to gather the best metrics on this time series data that I can in order to inform my choices on what I consider best practices. Do I need to get an IRB involved? My institution doesn't have an IRB.
I really wish there were high quality controlled experiments in education that I could rely upon. But the vast majority of the educational research is crap.
Think about, say, Khan Academy. I have a hard time imagining a world where interacting with millions of kids on a protocol is OK, but the second we take some basic metrics to try and figure out how well it works, it's bad. I'd arguably ethically one needs to perform a certain basic level of monitoring and optimization of any intervention.
Intent doesn't matter unless the data collection includes legally binding provisions to delete all collected data in the case of bankruptcy or acquisition. Otherwise this is just a pile of user data waiting to be scooped up by private equity and sold to third parties.
> In fact, there is very little money to be made by making online education more effective.
Right. So when the ed-tech company gets into financial trouble the data becomes an asset to be sold.
Educational systems fundamentally need to collect data on student learning progress. It is how education functions as a system. Lumping together the unethical (selling PII) and the ethical (continuous improvement) is very problematic.
Particularly:
> must not condition participation in any activity on a child disclosing more information than is reasonably necessary for the child to participate in that activity.
> are strictly limited in how they can use the personal information they collect from children. For example, operators of ed tech that collect personal information pursuant to school authorization may use such information only to provide the requested online education service.
They're not prohibiting the collection of data, they're strictly limiting what can be done with it.
I think this is good middle-ground privacy policy in general.
[0][PDF] https://www.ftc.gov/system/files/ftc_gov/pdf/Policy%20Statem...
That may prevent gathering data about, for instance, their performance on other related skills.
> are strictly limited in how they can use the personal information they collect from children. For example, operators of ed tech that collect personal information pursuant to school authorization may use such information only to provide the requested online education service
How about using their PII to link data sources together? That might be done for the improvement of related systems, rather than to provide the educational service used by the child. I mean it’s arguable, but this could create a real chilling effect.
The popular tool Lock Down Browser is more or less spyware that takes over your device, monitors every click, and watches every movement. And yet most colleges and universities can’t reliably detect cheating even with this tool.
... during the administration of an exam.
> and watches every movement.
This is an option most exams don't opt for.
> And yet most colleges and universities can’t reliably detect cheating even with this tool.
It does remove the massive temptation of using the computer you're sitting at to look "just one thing" up.
My big complaint about it is that it's garbage.
I would say the majority of tests are not well designed, and rely on rote memorization and recall. It's disheartening to see schools implement surveillance to compensate.
In a similar vein, after 9-11 the airlines loved the idea that they had to start checking the ID of every passenger to ensure it matched the name on the ticket. They said they supported this for "increased security" but the real reason the airlines liked the idea was that it eliminated ticket resale, which had been a major annoyance to the airlines' bottom line for decades.
Half of them clearly stated that they were collecting data and yet that school didn't have the kids sign any release from the parents. These are kids under 13 years old BTW. Essentially, the school decided for parents it was okay for random Chinese and other apps to spy on students without the parents permission.
there are no strong anonymity requirements. no discussion of merging student's profile with their google ad profile in N years when they're out of school, or defaulting to being logged in in the browser.
this reads like the ftc has no idea what the harms here are and is unwilling to come down on schools for requesting things that are dystopian, like eye tracking in exams.