From the instructor's point of view, there's no winning here -- there will always be a cutoff, and any position of it will lead to complaints. Even for an informal system, there's always a last person allowed in, and hence a cutoff.
If you have a good reason for an extension, you can still get it. But there's no reason anybody needs, specifically, the extra second after a formal deadline. If the deadline were moved to a second after midnight, you would try to submit two seconds after midnight and complain, and so on. (I mean, why do you think deadlines are at midnight in the first place? Because in-person collection would stop working at 5 PM, leading to complaints, so deadlines were extended to midnight to let people work at night.)
This is pointless slippage, and the resolution from the student side is so easy. Just pretend every deadline is 5 minutes earlier than it actually is, and you'll never have problems.
- Due dates are hard (usually midnight)
- Everyone gets 3 "emergency tokens", which give you a 24 hour extension on the assignment, no questions asked. Max 2 per assignment.
- On occasion, you could get an extension without using a token by speaking to the professor in advance, if you had extenuating circumstances.
It was a good compromise between meaningful deadlines and humanity. One particular advantage is that, if you missed the midnight deadline, you knew you were using a town anyway, so there was no pressure to sacrifice sleep to get the assignment in "on time".
Most people I knew were happy with it.
They gave me the freedom I needed to make micro-adjustments to my college/life balance and greatly lowered stress.
As a student, it was, well, nice. It was true that it was a bit harder to be motivated to finish a paper right away, but it also made the process far less stressful. When I turned in papers I thought they were the best they could be.
Of course, I'm sure there were at least some students in her classes who ignored everything until the end of the term or something, and were quite stressed out at the end...
This doesn't work if your class builds on earlier material (most STEM classes). As the professor, I need some feedback as to how well the class understands what I just taught.
Though I ended up with weird reviews. I got just as many students saying I was their best teacher as those saying worst teacher. I also had a surprising number of students try to turn into me the skeleton code I provided them claiming they did the assignment (they'd add like a line or two). Also had students say I was unavailable even though I lived in my office with my door open (students frequently came by outside office hours) and replied to emails between 9am and 1am. Learned that the complaints don't always match what you're doing and are hard to interpret.
Tldr: students will complain no matter what. Just be human. It's easier and less stressful for everyone involved.
As a CS professor, the one thing I could do that flipped things on their head was running an automatic submission and testing system for any assignment.
You're getting 0% or 100%, but you know at all times over roughly 2 weeks exactly what score you are getting. You can resubmit over and over. Yes, I'm going to run extra tests once the deadline passes to make sure no one is punking me, but it's rare that someone who gets 100% on the initial tests doesn't get 100% on the later tests.
And, for those of you who think that 0%/100% is a poor programming assignment scoring system--I'll ask you the same question I ask my students: "0%/100% is bad? So which of you are cool with partial credit when I miscalculate your grades?"
Is there really? In my faculty, doors were kept open. I entered lectures at 20% of time left and nobody would bat an eyelid.
Hard deadline system with some conveniences.
It was possible to upload the work multiple times, and for an hour or so after deathline, to select which of the uploaded copies would be the hand over.
This allowed for e.g. uploading some code that works, and later progressively optimized/improved versions, without fear of uploading something partially broken by the end.
I have seen deadlines wrap all the way around a 24 hour cycle in this way. 5 PM is bad because people want to work at night, so delay to midnight. Well, people aren’t finishing, so delay to 5 AM, since everybody would be asleep then. Well now people are blaming you for “forcing” them to do all nighters, so delay to 9 AM. Now people say that’s unreasonable, because submitting later in the workday wouldn’t hurt anyone. So delay to 5 PM.
In my undergraduate education, toxic behavior like this was common. One professor who was a particular asshole deliberately assigned onerous assignments to coincide with an assignment for another critical, difficult class that 75% of the students were taking. Academic hazing was pretty common in CS.
Why? Because empathy at scale works in ways which breeds hostility among groups and it creates an unhealthy incentive to compete over a teacher; it's not really reasonable to expect people to be immune to race or class or sex. Further, if the rules are strict then people need to know rather than _some_ people getting the discretionary treatment and others not.
Is it actually _better_, or does it just maximise specific metrics that you're able to easily quantify?
More to the point - do you want to _be_ that person?
I'm glad that my professors were human beings and not automatons - I probably wouldn't have done well otherwise (bureaucrats are not my idea of good role models).
For example, if you are arriving for a family dinner, if you are 10 minutes late, it's NBD, but if you are 3-4 hours late you have just pissed everyone off.
If you constantly ask for slack from your friends, to the point where it becomes common, your friend's internal & informal credit rating of you would go down, and they start reducing the amount of slack they gave you.
So a policy that reproduces the above would be: Deadline is 4pm, plus a 30 minute 'slack time' for extenuating circumstances (which everyone will put as 4:30). If a problem has come up, then come to me and we will talk about an extension. I won't give you many extensions although if you show a pattern of relying on them.
Imagine a real family that applies these deadlines. You walk in the door and your mother starts talking about applying for an 'extension' for your lateness.
Perhaps at the end of all of this you have the same outcome, that every family member feels the same way about you.
But you don't feel the same way about them in both cases, because in one, they act like robots, and in the other, they act like the imperfect people we all are. In one, they're strange threats, in another, they're lovable human beings.
I can't express strongly enough how nefarious this trend towards engineering all aspect of the human experience into 'process' is.
Theoretically, me having this conversation with you on HN is the same as real life. Practically, it's completely different and the latter would be far preferable if we could do it in the same amount of time, because we're not computers, we have real emotional and physical needs that are unsatisfied if everything is reduced to the most "optimal/efficient" interaction.
It feels like you treat others as threats. Your employer, your professor, the police etc, all threats. The better model is to create a society in which they are your equals - many countries have achieved that - not fence them off as some sort of 'different class of human' and then mitigate the damages.
Why then do some people demand cameras? Don’t they like the good old neighborhood discretion? Why is it so hard to feel for those who receive the shitty edge of discretion? Is it because the other side is so sweet?
Another side of discretion is that some judges are harsher in sentencing before vs after lunch. But such a human interface...
Police cameras are there to make sure power is not abused, they are not there to eliminate discretion.
There are actually multiple actors in the justice system that have a large degree of discretion, the police, the prosecutor and the judge. To me this indicates that your thesis is erroneous.
Discretion is a matter of trusting power, and as I've mentioned, discretion shrinks in front of a camera. When cameras provide the kind of ammunition to humiliate, sue, or pressure for job termination, then your discretion is compromised. Why did one forgive one youthful offender of marijuana possession and not the other? Was it race? These are potentially humiliating questions.
We can also contemplate what it means for employers to extend monitoring to social media, and whether that compromises your willingness to express your thinking in forums, and whether monitoring employees has anything to do with control and trust. Recently US border patrol agents were found discussing Latinos and other minorities in depraved ways. Does observation of their social media restrict their individual expression? Yes it does.
As to the fact that judges and other parties have discretion, well, I have at least forwarded one example of judges sentencing more harshly based on whether they've had lunch. That is part of discretion -- that some people get to enjoy the sweet fruits and others do not. This humiliating fact about judicial discretion would not have been known without observation.
It's a really weird comparison to me. Perhaps US profs are way more uptight in general? I'm not really sure how to respond to that, to be honest.
As far as I can tell this is simply a strawman. And even if it's not - the solution is to replace dickhead staff with reasonable human beings, not chain them with process so that everyone suffers.
The key here is, surveillance shouldn't be used against the students who are of low power, but instead against those who decide if the students succeed or not - the professors.
Trusting power is how one gets to significant asymmetry and gross abuses. Surveillance against those types is how you start to counterbalance "discretion".
https://www.indystar.com/story/news/education/2019/11/24/eri...
Like I said, attendance at my University was not mandatory. We weren't treated like children.
For many lectures I felt reading the book would be a better use of my time and so did others.
I think a decay-based system or an “emergency-token” system doesn’t discriminate between students, but still manages to provide humane leeway to get things done.
You mean a tech companies support as many anecdote has attested over the years upon Ycombinator with cries of trying to just speak with a human and that sad trend that if you want tech-support with a human it can boil down to knowing somebody who knows somebody who works there who can humanise the issue.
With that in mind, think we are already there alas. The aspect that you get emails from companies that are donotreply and to actually reply entails reading thru numerous fineprint to find an email that will automatically process you response.
Yes the days of drive-by support are well are truly here.
Speaking as a professor, there are plenty of severe and unfeeling professors out there. Thinking both as a student and as a professor, I'd much rather have everyone subject to the same inflexible rules, rather than to have some students favoured and some disfavoured based on what will inevitably be a subjective opinion—I'll be likely, even if subconsciously, to favour a student who has made a positive impression over one who has made a negative impression (when it is likely the latter who needs more help to get back on an equal footing!), or a student who asks on a good day than one who asks on a bad.
Then for the next two hours (typically until midnight) you can still submit for 95% of the points. I often also allow even later submissions, but with increasing penalty up to 50%.
My university was very reasonable though, they let me sit an exam 2 days late because I had to fly out for a job interview, they simply made me promise that I wouldn't ask my classmates about the exam. Funnily enough I actually ended up failing that paper anyway, but the opportunity was appreciated.
You're attending a lecture at a college. You're not in a private location. Traditionally, what the beacons were doing would be done by TAs taking attendance, class participation grading, or simply by having trivial pop quizzes during class. Adding bluetooth beacons just makes the process more efficient (and makes it easier for students to fool the system), but it doesn't change its essential nature.
Making lecture attendance a part of your grade may seem like bad policy or infantalizing to you, but that doesn't make it an invasion of privacy or social-credit policy.
Your tangential advocacy about "social understanding" missing the negative consequences of that approach. In particular that inconsistency of rule application creates a hole you can drive a truck through for all kinds of unfair and biased consequences (literally exactly what is brought to the table by a "feeling" application of the rules), and with little gained. While thoughtful empathic responses can be very helpful in a lot of situations, I don't think you've made a good case for this one.
It's true that unwritten social rules aren't 100% fair. They're not entirely consistent either. But the foremost one I was thinking about was the increased acceptance of human margin of error in accounting: humans, unlike computers, don't count down the the nanosecond. I've had assignments automatically marked as late because they only finished transferring over the network a second after the due time. A human would never say that's late. Could you tell me that I should be submitting earlier? Sure, fine, but I would still want to reserve the right to correct a final typo. Thankfully in all those cases I've been manually marked not late. But I don't think it's entirely beneficial for us humans to deal with a machine-type accounting of time for our own actions.
You don't know humans very well then. A human might say it's late even if you turned it in on time. Heck, a human might say you never turned it in.
> Could you tell me that I should be submitting earlier? Sure, fine, but I would still want to reserve the right to correct a final typo.
Hey, I would still want to reserve the right to correct a final typo from thirty years ago. I would still want to reserve the right to correct a final typo from a HN comment from weeks ago. It's nice wanting things. That doesn't make them good ideas.
A computer might try to make corrections right up until the last nanosecond, but a human might wisely recognize they need to get their corrections out of the way well ahead of time and stop looking for typeohs after they've handed in their assignment. ;-)
> But I don't think it's entirely beneficial for us humans to deal with a machine-type accounting of time for our own actions.
That's absolutely true, and in those cases you'd expect a different approach, regardless of whether there was a machine keeping time or not. I just don't think you've found such a case here.
I wouldn't suggest a measure where a prof was relying on a huge digital clock. I was trying to address a context that wasn't machine driven as you were advocating. There are simpler realities. A prof who will accept submissions until they leave for the day. For remote education, the physical mail thing is very real (or at least it used to be).
If neither party is benefiting from machine precision, there's no reason for the onus to be on the professor or administrator to make allowances for that imprecision. It actually makes a lot of sense for that onus to be on the student.
If you had a truly extenuating circumstance, you could always speak to the lecturer via email or during their office hours. Some had more relaxed policies than others, while others had various penalties for being late.
As far as I'm aware, the difference between 12:00:00 and 12:00:30 wasn't ever really a thing, as any reasonable system can expect a few minutes of delay owing to computer problems.
I had professors who would give their home address so you could get your assignment turned in on the calendar day and not get docked for it being late -- distinctly remember having to drive to the next town over to do that once.
Then again this was before the internets were all popular so they couldn't really expect the students to submit it by email.
Further, while there have been obvious benefits of technological advancement, they have in large part erased informal procedures and social understanding in favor of a rigid and unforgiving process. For example, my assignments are now largely due at midnight, at a very particular second: submit a nanosecond late at your own penalty. Whereas in the past the common practice of most professors would be along the lines of "get it in my mailbox by 4 p.m.", or something like that, but with the informal understanding that at 4:00:00.0...1, the door of the mailbox would not close and chop off your hand. And exceptions to rules were easier to get before the arbiter of the rules became a severe and unfeeling one.