Before covid, "remote proctoring" tools were a niche product, invasive tools that spied on students who needed to take high-stakes tests but couldn't get to campus or a satellite test-taking room. But the lockdown meant that *all* students found themselves in this position.
1/
(If you'd like an unrolled version of this thread to read or share, here's a link to it on pluralistic.net, my surveillance-free, ad-free, tracker-free blog:)
This could have prompted educators to reconsider the use of high-stakes tests. After all, high-stakes testing has well-understood limitations in pedagogy, and organizes education around a highly artificial ritual completely unlike the rest of scholarly *and* industrial life.
3/
It's not like anyone does a *job* where you are prohibited from consulting reference texts or collaborating with your colleagues (if you have an colleague who does this, you should probably ask to be transferred to another team).
4/
While in the academy, neither scholars nor researchers work without collaboration or access to references. It's not clear what, exactly, a high-stakes test measures, apart from your ability to engage in the useless, non-transferrable skill of sitting a high-stakes test.
5/
But rather than rethinking assessment, educational institutions doubled down on remote proctoring, throwing stupendous sums at companies that made outrageous promises about their ability to automatically detect cheating with "AI."
6/
While this threw every student into a meat-grinder of opaque algorithmic cruelty, not every student suffered equally. In "Rejecting Test Surveillance in Higher Education," @GeorgetownLaw's @LAMBarrett describes the unequal and disproportionate harms.
While Barrett's paper is long and thoughtful, the introduction tells five stories, smartly making the case that if you are already marginalized, remote proctoring hurts you more than your fellows, magnifying your existing disadvantages.
8/
Jazi: "a first-generation student" and caretaker (thanks to covid), who "was flagged for 'suspicious noise'" (small children) "by the software recording her take an exam in her family’s home," who "emptied her bank account to pay for on-campus housing."
9/
Ashley, who "dissolved into tears" when "a remote proctor with access to her computer kept moving her cursor...distracting her and exacerbating her anxiety about her performance."
10/
Audrey, "a transgender student at the Georgia Institute of Technology" who was "humiliated" when a remote proctor demanded their professor to confirm Audrey’s identity thanks to a photo ID that didn’t match their gender presentation.
11/
Tracy, a OSU grad with ADHD who could only use "'half her brain" on her remotely proctored exam, using the other half to police "the minor fidgeting that is a side effect of her condition, as she was afraid the software would characterize her movements as suspect."
12/
Ahmed, "an Arab-American law school graduate, estimated that he tried 75 times to get the remote proctoring software he would be required to use on the bar exam to recognize his face. It never did."
13/
If you are poor, if you have a disability, if you struggle with your mental health, if you are racialized, if you are gender nonconforming, then remote proctoring tools punished you for who you were, far more than it did your peers.
14/
Remote proctoring companies were incredibly successful covid profiteers, growing by leaps and bounds during the pandemic. As they grew, so did the harms they imposed on the public.
15/
The companies responded to bad news about their products by deploying their covid riches to threaten their critics. @Proctorio went after educators like @Linkletter with a baseless - but ruinously expensive - copyright claim to silence his criticism.
But Proctorio didn't limit its harassment to adults who could take care of themselves - the CEO *personally doxed a student* who criticized his product on Reddit.
Proctorio isn't an outlier. The entire sector runs on bad faith, bullying and legal harassment. EFF's massive roundup of the lies and frauds of the remote proctoring sector is a good place to refresh your memory.
Be sure to read to the end, where Jason Kelley and Haley Amster document how, after apologizing and promising changes, the leading remote proctoring companies are *still* wreaking havoc.
19/
After all, these are companies that claim they can automatically rank whether students are "high integrity" or "low integrity" and produce "suspicion rankings" based on data analysis. This isn't computer science, it's digital phrenology.
20/
Barrett ends well: "Recommendations For Use of Remote Proctoring Software: Don’t." Instead, she sets out a of alternatives to high-stakes testing, including "long papers graded at multiple stages," "short response papers" and "presentations with a peer critique component."
21/
If you must use timed, remote exams, change parameters to make gaming harder: allow open books, randomize question orders, and encourage collaboration.
22/
If you need to hire more TAs to help grade these more pedagogically useful assessments, look to the savings from ditching remote proctoring tools, like the $500k/year that some schools are paying Proctorio.
23/
Above all, Barrett says, don't normalize this. This was a mistake, something that compounded the pain of the lockdown. This is the moment to reassess our approach: to reverse the alarming trend of increasing commercial surveillance of our students.
I've been paying close attention to @RepThomasMassie during the #ACCESSAct markup and I can't figure out his point. He correctly observes that proprietary standards are anticompetitive, but opposes the gold standard for open standards, namely, an IPR policy requiring licensing
@RepThomasMassie has described himself as a software developer, but it really feels like he is way, way out of his depth on standardization. Has he ever participated in an SDO. Not being able to distinguish between "interop" and "common vuln" is a pretty tyro error.
It's stuff like this that makes people assume that lawmakers are incapable of understanding - and thus regulating - technology. @RepThomasMassie really needs to get up to speed on how standards work.
In the #ACCESSAct hearing, @RepThomasMassie called the shared vulnerabilities in large-scale hacks as stemming from "interoperability." That's factually wrong. They have "shared dependencies" (use the same code/modules). This isn't the same thing as "interoperability."
Then @RepThomasMassie correctly warned that when firms get to define standards to their proprietary advantage, it produces monopoly power. However, #ACCESSAct provides for OPEN standards, developed independently of large firms.
The problem of proprietary advantage through capture of standards is well-understood and the #ACCESSAct takes account of it.