Food for thought: I like this idea of banning technology in the classroom (excluding accommodations) and putting students on a track of hand-writing assignments, so much so that I'm considering...
Food for thought: I like this idea of banning technology in the classroom (excluding accommodations) and putting students on a track of hand-writing assignments, so much so that I'm considering implementing this for my 400-level course next semester. This is going to add a lot of work to my plate, but I think removing ways for students be distracted in class, no matter how good intentioned they are, and having to actually write out their thoughts (or even copying it down from an LLM) should hopefully improve engagement with the course.
Outside of disability and English-second language individuals, what are some potentials pitfalls of this approach? Would you have liked this as a student, if you were one that used your computer to take notes?
I type literally 10x faster than I can write, and half of my professors could not read my handwriting in early college. Im sure part of that is being out of practice, but still, you had been my...
I type literally 10x faster than I can write, and half of my professors could not read my handwriting in early college. Im sure part of that is being out of practice, but still, you had been my professor and had done that, I probably would have bought a typewriter or taken my laptops wifi chip out of my laptop in front of you. I dont use AI, hope to never have to, but a computer is the only way I could have ever hoped to keep up with my notes. In some classes with fast speaking professors, even that wasnt enough, and I bought a dedicated voice recorder for those classes.
I certainly understand that through-put is much higher with typing for most people. The organization of the keyboard and the nature of typing is pretty much always going to beat out handwriting....
I certainly understand that through-put is much higher with typing for most people. The organization of the keyboard and the nature of typing is pretty much always going to beat out handwriting.
But when it comes to learning, going fast is not really all it's cracked up to be. It's generally better for us to slow down the learning experience because it helps encode the information more efficiently. A recent meta-analysis backs this up - handwriting notes tends to improve retention and leads to better academic outcomes.
For what it is worth, I hate this line of logic and it's caused me no end of friction. People learn in different ways, I did plenty fine typing out my stream of consciousness notes and despised...
For what it is worth, I hate this line of logic and it's caused me no end of friction. People learn in different ways, I did plenty fine typing out my stream of consciousness notes and despised the classes that forced me to "slow down" because their definition of slow was much much slower than mine.
I understand the good intentions here, but I would outright refuse to take a class that wouldn't let me type in this day and age. I also think it's doing a disservice to the students to not prepare them to use the tools they'll actually be using in real environments.
This feels like the kinda of academic optimization that's far better in ideals than reality.
That's fair, and I would be upfront about my policy at the start of class so students can drop if they want to. Ultimately, because people learn in different ways, no matter the approach and...
That's fair, and I would be upfront about my policy at the start of class so students can drop if they want to. Ultimately, because people learn in different ways, no matter the approach and pedagogical techniques I use, some students will benefit while others will falter. Such is the reality of not doing individualized instruction. So I'm going to go where the evidence leads me on best practices to help the most amount of students.
I don't necessarily disagree that we should prepare students to use the tools that they'll be using in a job, but I view that as being subordinate to developing the foundational skills that will maximize the utility of those tools. And in my experience as an instructor, many students desperately need to practice working through their thought process and developing arguments. On top of that, I don't think I'm doing students a disservice if I don't let them type notes in my class - they already have plenty of experience with using a computer.
That's fine, but I don't agree your solution will actually lead to that, and it seems more like an arbitrary hurdle that's got research from some ideal setting rather than well tested. I had a...
And in my experience as an instructor, many students desperately need to practice working through their thought process and developing arguments.
That's fine, but I don't agree your solution will actually lead to that, and it seems more like an arbitrary hurdle that's got research from some ideal setting rather than well tested.
I had a teacher in high school who said we should have to write only on unlined paper because that's what would be expected of us going into college. That was NEVER true, never helped, and made me drop out of the class because it unfairly punished people like me with bad handwriting as I'd get docked points for not following where the lines should have been (we were supposed to put lined paper underneath the white sheet to adhere).
She was a good teacher, and instead I chose to drop because of the arbitrary restriction. I don't think yours is objectively as bad, but I doubt you're going to see the results you want. I wish you luck with it though.
That certainly is a pretty annoying and harrowing rule to have. I wonder if it may have colored your perception of having to write things by hand? It certainly would make me aversive to the idea....
That certainly is a pretty annoying and harrowing rule to have. I wonder if it may have colored your perception of having to write things by hand? It certainly would make me aversive to the idea.
Hopefully it does goes well for my students. I have no problem going back on it if causes more issues than it's worth. Thanks for the discussion.
Its not that typing is faster, its that my writing is slow enough, and professors fast enough, that handwriting in early college often led to me only encoding 2/3rds of the information at all,...
Its not that typing is faster, its that my writing is slow enough, and professors fast enough, that handwriting in early college often led to me only encoding 2/3rds of the information at all, leading me to doing horribly on tests if I wasnt able to use recording devices or borrow somone elses notes.
In my degree only half of the classes had textbooks, and only half of those that did ever told us what the material we should read was ahead of time. When provided, I much preferred reading the...
In my degree only half of the classes had textbooks, and only half of those that did ever told us what the material we should read was ahead of time. When provided, I much preferred reading the textbooks to the lectures and would read them both before and after, yes.
OK thanks for answering. I don't require a textbook, but I do suggest one and I emphasize to my students that they should be reading about the material before coming to class that way when they...
OK thanks for answering.
I don't require a textbook, but I do suggest one and I emphasize to my students that they should be reading about the material before coming to class that way when they hear new terms they aren't completely lost. I really don't want to make textbooks mandatory because of how expensive they are, but I think it's very beneficial to read about these concepts (barring some sort of reading/learning/visual disability).
Yup, textbooks make everything easier, but I hear you WRT the expense. I had been sailing the high seas for quite a bit before college, and during college I kinda became the local 'pdf dealer' in...
Yup, textbooks make everything easier, but I hear you WRT the expense. I had been sailing the high seas for quite a bit before college, and during college I kinda became the local 'pdf dealer' in my program. I know over 50 people who got all of their textbooks from me.
To be fair, that's probably more of a professor issue than a you issue. Some professors seem hell bent on covering too much imo. It feels counter intuitive, but in my experience it's more...
To be fair, that's probably more of a professor issue than a you issue. Some professors seem hell bent on covering too much imo. It feels counter intuitive, but in my experience it's more effective to cover less in a lecture, take it slow, and provide additional out of class resources (and definitely not just out of class readings that everyone obviously actually reads) to fill in gaps.
Yeah ... I don't know how much I trust that meta-analysis. For context: I'm used to seeing meta-analyses use a robust framework for rejecting studies that are poorly designed, or are clearly...
[Studies] had to meet six screening criteria for inclusion in the meta-analyses. Published and unpublished studies meeting these criteria were included. First, the studies had to compare how recording typed (e.g., laptop, desktop computer, tablet) versus handwritten lecture notes affected note-taking and/or achievement outcomes. Acceptable achievement outcomes were scores on immediate or delayed lecture-related post-tests, course exam or quiz grades, and final course grades. Second, participants needed to be college students. Third, the instructional materials must have been either a live classroom lecture or a pre-recorded video lecture in a laboratory setting. Studies including remote learners were not included. Fourth, only experimental and quasi-experimental studies were included. No survey-based or observational studies were included. Fifth, lectures needed to pertain to college-level topics; studies whose instruction focused on list learning or other non-meaningful learning outcomes (e.g., Aragón-Mendizábal et al., 2016) were excluded. Sixth, studies needed to have measured a direct effect between hand-written and typed note-taking and provided adequate information and data about the measured effects, such as test statistics that compare handwritten versus typed note-taking or descriptive statistics that included information such as sample sizes, mean scores, and standard deviations for both note-medium groups.
For context: I'm used to seeing meta-analyses use a robust framework for rejecting studies that are poorly designed, or are clearly biased. The fact that they also included non-peer reviewed studies is kind of maddening, too. For example, why were remote courses removed from the analysis? They were fine with pre-recorded lectures in a lab, why not at home? The authors even note that inclusion bias could be screwing up several other meta-analyses which have shown no result, instead of demonstrating a superior medium (see Findings from Previous Meta‑analyses).
I don't have the brain space atm to read through this paper (How to interpret discrepancies in empirical results from educational intervention studies.) (pre-print here), but IMO at a skim it confirms my biases at least: that educational intervention studies are extremely difficult to design, they get blown out of proportion by people with an agenda, and that we basically know nothing and should stop taking away tools from students.
(edit) Sorry, just clarifying my point, because I'm pretty scatter brained atm. People have been complaining about "kids these days" since time immemorial. Equally, seemingly everyone has an opinion about how kids should be raised, or what is the morally correct way of doing things. I think it's all bupkis and pseudoscience, and that you'd find a dramatically larger effect size on student performance with other, less contentious interventions: pay teachers more, decrease class sizes, change incentives towards learning instead of rote memorization/cheating.
(and for context, I can handwrite ridiculously fast anyways, and preferred switching between longhand and typed notes in university, based on the class. Calculus, discrete math, etc. got the wacom tablet, whereas psychology, linguistics, etc. got the keyboard. Depriving students of tools only makes them less capable humans overall)
I'm not too interested in litigating all of the research on this because I'm not that committed to the idea that things must be handwritten in order to be learned. I know many people encode...
I'm not too interested in litigating all of the research on this because I'm not that committed to the idea that things must be handwritten in order to be learned. I know many people encode information from typing and by reviewing the notes they typed. There is no one-size fits all approach to note-taking and education that could be implemented in a classroom. The main reason I subscribe to the idea of writing things down as a good technique for students is because my pet hypothesis is that forcing students to slow down is better for retention (which I seem to recall is backed by some study, but I don't have the time to dig for it at the moment), and I think that most students aren't taking that time to slow down. There are many perverse incentives for why they don't slow down, such as overloading their schedules with too many classes and not having enough time to dedicate to studying while balancing personal life and work.
Regarding your last point about less contentious interventions - I don't think the first two are less contentious, otherwise we would simply do those, no? Yes, I would love to have less students so that I can work with them more directly and paying teachers more would hopefully increase retention of good teachers and encourage more people to join the profession, but we pretty clearly don't want to do that as a society right now and so I need to think about other ways of trying to improve my pedagogy given the system I'm in.
No worries. I'll bow out; I was concerned that people in this thread were reading a study, thinking it reflected reality, and were attempting to act in a research-driven manner. If we're all kinda...
No worries. I'll bow out; I was concerned that people in this thread were reading a study, thinking it reflected reality, and were attempting to act in a research-driven manner. If we're all kinda winging it, that's fine. Rigor isn't necessary anywhere, it's just annoying to see people wag scientific papers around to prove their point, when vibes were always a perfectly reasonable way to ground your decisions.
I won't dig into your perspective; I disagree with it, but I'm doing so based on different grounds than you (i.e. it seems that the field can't agree on anything, so claims can't be made either way), so it's not terribly productive to discuss it further.
Regarding your last point about less contentious interventions - I don't think the first two are less contentious, otherwise we would simply do those, no? Yes, I would love to have less students so that I can work with them more directly and paying teachers more would hopefully increase retention of good teachers and encourage more people to join the profession, but we pretty clearly don't want to do that as a society right now and so I need to think about other ways of trying to improve my pedagogy given the system I'm in.
Yep, fair enough. I wasn't trying to suggest that you do anything different. Mostly, I'm suggesting that Business Insider stops posting articles which shape a narrative that actually laptops are the problem w.r.t. academic achievement. Not systemic issues in the school system.
(life issues are draining my brain at the moment, and I apologize for any bluntness in the above comment. I'd normally give it a few more editing passes, but I have several issues to attend to. Good luck with your class; I'm glad you're trying to do what's best for your students)
Truly, thanks for sharing your thoughts. I do agree that people will drop papers all the time as justification without much further thought - the main reason I wanted to share the meta analysis...
Truly, thanks for sharing your thoughts. I do agree that people will drop papers all the time as justification without much further thought - the main reason I wanted to share the meta analysis was that it's not a crazy idea to encourage students to shift to handwriting (but as you pointed out, the evidence really is murky, thanks for sharing that article in your previous response. I intend on reading it over spring break). I just didn't want to get into an intense discussion about what constitutes valid evidence for a meta-analysis because that can be quite a quagmire.
Certainly Business Insider, and other publications, will be looking for simple explanations for complex issues. These ultimately are caused by systemic issues, and I am totally sympathetic with you on that. If I was the education tzar and had a lot of power to change the education system, it would look drastically different than what it is right now. One of those things would not be the tzar and actually having a panel of informed people to help make decisions.
Handwritten notes being shown to improve outcomes on a population level does not entail that it's an improvement in every individual circumstance. You can't wave around results like that as though...
Handwritten notes being shown to improve outcomes on a population level does not entail that it's an improvement in every individual circumstance. You can't wave around results like that as though they universally apply to individuals. That's not how science works. And I say this as someone for whom handwriting notes does improve things, personally.
I didn't claim that it universally applies - I'm aware there is no one size fits all approach to learning via note-taking as I've mentioned in other comments on this post. And in fact, I...
I didn't claim that it universally applies - I'm aware there is no one size fits all approach to learning via note-taking as I've mentioned in other comments on this post. And in fact, I acknowledge that in the comment you're replying to by explicitly typing out "generally" and "tends to". However, it seems from the data I've read that more people benefit from handwriting notes than other methods.
In a similar vein, not everyone sees benefits from exercise when it comes to the numerous health outcomes we associate with it. However, I don't think it's faulty to suggest that generally exercising is a good idea.
I don't think that citing that meta-analysis in response to someone saying they struggle with handwriting notes and do better when they can type amounts to just observing trends, but is instead...
I don't think that citing that meta-analysis in response to someone saying they struggle with handwriting notes and do better when they can type amounts to just observing trends, but is instead giving advice to an individual who has already expressed their lived experience about how these methods affect their note-taking. Even if the research on this were rock-solid (and I'm not convinced it is), I think it may be worth imagining how frustrating a response that is so dismissive of your own experience with notetaking and your own brain would be, even if they pay lip-service to there being individual differences. I have friends with chronic illness who have the same experience with people constantly insisting they should try yoga even after they've mentioned that yoga would actually exacerbate their particular disabilities, so perhaps exercise isn't the worst comparison. It's tone-deaf, imo.
As other comment in this thread noted, I think this is a feature not a bug. But I do understand my frustration : my writing style is messy and iterative ; there is a cost in writing stuff down...
I type literally 10x faster than I can write,
As other comment in this thread noted, I think this is a feature not a bug.
But I do understand my frustration : my writing style is messy and iterative ; there is a cost in writing stuff down over and over again.
Isn't that the solution? I'm not sure there's much value in being a stenographer. Notes should be more condensed than natural human speech, which is not particularly information dense by default....
I bought a dedicated voice recorder for those classes.
Isn't that the solution? I'm not sure there's much value in being a stenographer. Notes should be more condensed than natural human speech, which is not particularly information dense by default.
Of course, professors can refuse to allow recorders, but they can refuse to allow anything - I once had a professor refuse to allow us to write anything down.
Personally, I am totally fine with student's voice recording lectures. I also agree with what you're saying about notes - you shouldn't be handwriting every single thing that I'm saying or that...
Personally, I am totally fine with student's voice recording lectures. I also agree with what you're saying about notes - you shouldn't be handwriting every single thing that I'm saying or that appears on my lecture slides. Take out main points and use those as a foundation for understanding the other material.
That really sucks that your professor wouldn't allow you to write anything - what was their rationale?
Taking notes inherently implies a structure where there's a speaker and listeners. That professor wanted a more communal form of teaching. It was a pretty niche class - a class on gnosticism and...
Taking notes inherently implies a structure where there's a speaker and listeners. That professor wanted a more communal form of teaching.
It was a pretty niche class - a class on gnosticism and other early esoteric christians sects. There were like 8 students, so we just sat in a circle and it was expected that if we didn't understand anything immediately, rather than try to take notes, we would simply ask for clarification.
I didn't mind, really. Was a good experience, if anything.
Ah OK - yeah I can definitely see that working. There wasn't an explicit rule about that in my graduate level insect physiology course, but the instructor set it up in a way similar to what you're...
Ah OK - yeah I can definitely see that working. There wasn't an explicit rule about that in my graduate level insect physiology course, but the instructor set it up in a way similar to what you're describing where taking notes didn't really make sense. I think that works really well for small class sizes.
Isn't it a problem only when writing absolutely everything ? Like doing a transcript ? I can only speaks from my point of view, but I never encountered a class where I needed to write everything....
Isn't it a problem only when writing absolutely everything ? Like doing a transcript ?
I can only speaks from my point of view, but I never encountered a class where I needed to write everything.
Whenever I handwrote something for my studies, it was always by summarize and rephrase what I've listen to, and choosing actively what was important to take note of, and what was optional to note.
And I don't remember one time re-reading my notes in details. It was a tool to help my brain re-connect information together.
The act alone of handwriting became a good way to remember the information I've just learned for way longer.
My only issue with AI note-taking is that physically handwriting the notes is better for retention - beyond that I would not care at all if a student was using a voice-to-text dictation as long as...
My only issue with AI note-taking is that physically handwriting the notes is better for retention - beyond that I would not care at all if a student was using a voice-to-text dictation as long as it helps them better understand the material.
I was a TA for awhile, and handwriting is a huge barrier to being able to read and evaluate a student's work. I think there are definitely advantages to doing assignments by hand in class - it...
I was a TA for awhile, and handwriting is a huge barrier to being able to read and evaluate a student's work. I think there are definitely advantages to doing assignments by hand in class - it forces students to complete assignments there and to potentially discuss course content with peers - but it was so miserable on the grading front. I would not do it if you have more than ~30 students at the most, unless you're cool with tormenting yourself and/or your TAs!
I had a few Francophone students at my Anglophone university, and I do think they benefitted from being able to practice writing by hand, especially if they had time to work with peers - having handwritten assignments slows everybody down, and having a more group-focused context can allow them to discuss ideas and support each other. So having ESL students isn't necessarily going to be a problem, depending on how the course is structured and how much they can work with other students. If they are all working independently, that may add additional barriers, but it might be worth chatting with those students to get their perspectives.
The class I'm planning on implementing this in is ~45 students. I can't keep up with grading a weekly assignment, so I'm thinking of maybe having them submit something every 4 weeks around exam...
The class I'm planning on implementing this in is ~45 students. I can't keep up with grading a weekly assignment, so I'm thinking of maybe having them submit something every 4 weeks around exam time. Hopefully that can give me a quick turn-around in terms of grading and getting it back to them without them having to rely on it for studying.
You can always test it out and see how it goes! I think it's great that you're thinking about how to achieve good learning outcomes for your students, and I think trial and error can be helpful to...
You can always test it out and see how it goes! I think it's great that you're thinking about how to achieve good learning outcomes for your students, and I think trial and error can be helpful to refine your vision. Good luck!
Doctors are notorious for having bad handwriting and have been for a really long time. I think legibility has more to do with how we’re taught to write.
half of my professors could not read my handwriting in early college. Im sure part of that is being out of practice
Doctors are notorious for having bad handwriting and have been for a really long time. I think legibility has more to do with how we’re taught to write.
My problem here is that we shouldn't be optimizing our education system for the top 10% of performers if it means worse outcomes for the bottom 50%. Having a capable, educated workforce benefits...
My problem here is that we shouldn't be optimizing our education system for the top 10% of performers if it means worse outcomes for the bottom 50%. Having a capable, educated workforce benefits everyone.
The compromise I see here is an extension of something we've already done for a long time: putting high performing students in accelerated education programs. A student who demonstrates above average proficiency and self-discipline could be placed in classes where assistive technology is more readily available.
I agree with you on this - we need to try and reach and teach the majority of people. Ultimately, we need more individualized instruction plans based on where the students are at. I don't think we...
I agree with you on this - we need to try and reach and teach the majority of people. Ultimately, we need more individualized instruction plans based on where the students are at. I don't think we can do it truly on an individual-by-individual basis (although that would be ideal), but perhaps having several streams of course concepts/practice extending and applying the idea of AP classes in high school - at all levels of education.
Even when I was in practice my handwriting was so bad and so slow compared to my typing speed. It was such a breath of fresh air to finally swap over to typing out papers and such. It sucks so...
Even when I was in practice my handwriting was so bad and so slow compared to my typing speed. It was such a breath of fresh air to finally swap over to typing out papers and such. It sucks so much now that its becoming more of a liability to teachers, but I can't imagine being forced to going back to hand writing out a 10 page paper, I'd cry.
It might work, depending on the type of class and the subject matter. However, I think you would need to make some accommodations for the change in approach. I agree with the comments about...
It might work, depending on the type of class and the subject matter.
However, I think you would need to make some accommodations for the change in approach. I agree with the comments about handwriting being slower and more difficult to read.
Some ideas:
If you are only focused on note taking, since some people will be slower than others you can adjust the content so there is less to write, or provide some type of handout so they have some notes already.
If it is an assignment, I don't think techniques like having a minimum word count would work. A requirement that is inherently easier to do on a computer can end up distracting from the objective of the assignment when asked on a handwritten assignment.
If you are going to read their handwriting, it would be good to increase the chances that their handwriting is as neat as they can make it. For example, some people will write neater when they have more time to write. Also, consider an "intro to note taking" session that can include topics like useful shorthand so ppl dnt thnk thy hv 2 fully spell out evry wrd.
Thanks for sharing these ideas! I like the idea of doing a short crash-course on note-taking in my introduction/go over the syllabus class. I certainly would not have a minimum word count for...
Thanks for sharing these ideas! I like the idea of doing a short crash-course on note-taking in my introduction/go over the syllabus class.
I certainly would not have a minimum word count for written assignments. I would probably focus on having some really clear rubrics that highlight what all they have to do in terms of addressing questions or engaging in discussion. Fortunately, I have all summer to think about and implement these changes.
Off the cuff, I would speculate this works well in a lot of classes but would potentially fall completely flat with a lot of CS related courses. Anything where the topic directly involves a...
Off the cuff, I would speculate this works well in a lot of classes but would potentially fall completely flat with a lot of CS related courses. Anything where the topic directly involves a computer or electronic device. There are probably variants of the notebook-only approach that would work for that as well, but might be more difficult to figure out and for the students to adjust to.
I never took university-level classes, but I think I would probably have hated it (like the student comments from the article) but ultimately been grateful that it was done that way.
You can absolutely do hand-written exams and assignments for most CS classes, and many of mine did, especially in the beginning of the degree. I think it was a good choice and helpful in the long...
You can absolutely do hand-written exams and assignments for most CS classes, and many of mine did, especially in the beginning of the degree. I think it was a good choice and helpful in the long run. A bit tedious? Sure, but it really forces you to learn programming logic and ingrain a deeper understanding of the topics in your mind. A lot of classes (like data structures) don't necessarily need actual programming right away either.
My favorite CS professor started off his intro class with extolling the virtues of hand writing notes, and he was absolutely right. The few of us that took this to heart ended up doing very well throughout the CS program, and generally seemed to be above average in our understanding (though tbf there always seemed to be a pretty bimodal split this way, especially when it came to the weed out classes). In addition, using a laptop during class was just far too distracting for me, so it was extremely beneficial in regards to forcing me to pay attention (or at least more attention than I would've).
Every single CS class I had used handwritten exams, and they were weighted to be the majority of the grade. At the freshman level, this often involved "here's a printed Javadoc" or "I hope you...
Every single CS class I had used handwritten exams, and they were weighted to be the majority of the grade.
At the freshman level, this often involved "here's a printed Javadoc" or "I hope you memorized C functions" and having to write out a program. In higher level classes, which quickly become about theory and not writing code, you end up doing things like labelling states of registers on a toy CPU based on assembly instructions, or drawing diagrams of how a red-black tree is changed at each step of an algorithm.
Honestly, programming is a prereq to start a CS degree. It's not what the major is about, and if you're not already decent at it, the limited classes that focus on it aren't going to do much to remedy that.
There are certainly environments where you need a computer or some sort of device - a programming course for instance would require it and I would not expect students to do otherwise. But, maybe...
There are certainly environments where you need a computer or some sort of device - a programming course for instance would require it and I would not expect students to do otherwise. But, maybe in the recitation where you are talking about programming concepts you wouldn't necessarily need to have a computer.
I was a CS major and the classes which required a computer typically took place in the lab, where there were desktop computers available for everyone. Towards the end of my degree (2018) there...
I was a CS major and the classes which required a computer typically took place in the lab, where there were desktop computers available for everyone.
Towards the end of my degree (2018) there were maybe 1-2 times that we ended up in a lab without computers because of some kinda scheduling mistake.
Its actually kinda weird to me that students these days almost always have their personal laptops in the lecture hall. Thinking about it now, I have tons of hand written notes from my CS major, I didn’t use my laptop during lectures often. I relied on my laptop mostly for homework, because although I could use the lab computers for my assignments, I liked my programming setup better and doing homework in my apartment beats having to stay on campus.
When I grew up there were no electronics in class. The best students could feel some level of exhaustion in the muscles of their hand and wrists. It wasn't anything serious as far as I know. I had...
When I grew up there were no electronics in class. The best students could feel some level of exhaustion in the muscles of their hand and wrists. It wasn't anything serious as far as I know. I had awful handwriting and sometimes couldn't read what I wrote. So lookout for those kids, teach them how to write correctly. And get adequate stationery for the left-handed like me. Other than that I think it's a great move.
I'm really thinking about high school and below here. Film school didn't require much note taking from me. I'm not sure how that would work for other majors.
Even as a student who does perform better when I limit myself to handwriting my notes and don't bring my laptop to class as a potential distraction, I don't think wholly banning them is...
Outside of disability and English-second language individuals, what are some potentials pitfalls of this approach? Would you have liked this as a student, if you were one that used your computer to take notes?
Even as a student who does perform better when I limit myself to handwriting my notes and don't bring my laptop to class as a potential distraction, I don't think wholly banning them is acceptable. Firstly, you cannot ignore disability and ESL as these will be hugely salient in almost any class you're in. Secondly, even people without a disability may have valid reasons they prefer to take notes on a digital device -- much like how I use subtitles despite not being hard of hearing. Just because they're a disability aid doesn't mean they can't often be helpful to people who aren't disabled too.
Thirdly, frankly, in a 400-course it feels incredibly patronizing and like being treated like a child. If it's a seminar and is more discussion focused anyway, I can see a no-laptops policy making sense (although even then, that's disruptive to those who save the readings as digital pdfs rather than printing them out, which was common in my field and has similar concerns when it comes to disability), but for a lecture or similar course where you're expected to take notes, it comes across as treating students like little children who are only allowed to do what you trust them to do. I remember when I first attended a college course as part of post-secondary enrollment, I was in awe of the fact that I was allowed to pull my laptop out when needed without the teachers acting like cops about it, because that was all I knew from the strict rules in high school. Being treated like an adult who could make their own choices about how they learned made me respect my professors more, and I eventually came to my own conclusions about what note-taking methods worked best for me. If I had been forced to take "analogue" notes by my professors, I would have resented it and never learned the ways that they work better for me personally, and/or would have avoided taking notes at all out of spite. And I'm a student whose disability means I benefit from taking handwritten notes. Imagine how much worse this feeling of overbearing patronization would be when compounded by a disability that makes it harder with handwritten notes.
Regarding your first point - I never claimed that I would prevent students that have a disability or ESL from using these devices? I'm aware that their learning may be greatly impaired if they're...
Regarding your first point - I never claimed that I would prevent students that have a disability or ESL from using these devices? I'm aware that their learning may be greatly impaired if they're not able to use electronic devices. Me phrasing my sentence that way is to acknowledge that there would be no question about them bringing it in.
That is a fair point regarding the rest of what you wrote. I imagine that if I do implement this rule some students would build up some resentment. Some students also, as noted in this article, might develop an appreciation and get more out of the class. No matter the expectations and interventions in the classroom, some students will benefit from a given pedagogical technique while others might suffer from it.
I think in principle you want to accommodate those who are disabled or have other barriers like ESL involved, but your descriptions in other comments of how you'd plan to present and implement...
I think in principle you want to accommodate those who are disabled or have other barriers like ESL involved, but your descriptions in other comments of how you'd plan to present and implement this policy indicate that this would not be free of negative effects on some of these people in practice. How do these people get accomodation so that they're exceptions to this policy? Do they need to go through your university's accomodations office? Do they need to personally ask you and reveal potentially private information about their disability? Even if these things are not the case and you grant accommodations to anyone no questions asked, these are barriers that accomodations typically require and the prospect of them existing here will probably discourage some subset of students who need them from asking for accomodations at all -- they'll either suffer through the course as best they can and perform worse than they otherwise could, or they'll drop your course and lose out of the opportunity to become engaged and learn what you're teaching in the first place. This is going to be especially true for those who don't have a formal diagnosis, as they'll have more experience being denied accommodations (or less experience fighting for them) than other students. ESL students are in a different circumstance but they have similar pressures to not ask for accommodations when a professor places a rule like this in the syllabus, even if accommodations are available. The existence of the policy itself will discourage people from asking for "exceptions" as accommodations. You simply cannot make a blanket policy like this without collateral effects on some of the people. If you choose to implement this policy, you need to be well aware that you will be denying access to some disabled people, even if that's not your intention. There is no version of this policy that won't affect some of those people, and implementing this policy entails you believe its benefits are worth that.
It's true that different pedagogical styles help and harm different students differently. But I don't think it's remotely clear that the benefits of this one for some students outweigh the potential harms for others. And even if they're roughly equal, why would you choose the policy that demonstrates a paternalistic lack of faith in your own students rather than one that gives them even a small amount of freedom to discover for themselves what works best for them when it comes to how to best study and take notes? I would never have learned that I personally do better taking handwritten notes without doing different things for different courses (sometimes on different days in the same course) and discovering this through experience. I, like your students, am an adult who can make my own choices even if they result in me not maximizing how much information I retain, and the freedom to make my own choices rather than being forced to take notes in the style my professor believes is better based on some meta-analysis they read is part of being respected as an adult who can learn and engage with the material rather than as a naughty child who must be reigned in so that they aren't disruptive in class. I would chafe at that bit and I don't believe that the positive effects of handwritten notes would remotely equal the negative effects of me resenting you in terms of my engagement with your course and the material. And frankly I'm probably the best case scenario, as I'm a generally highly engaged student who does benefit from handwriting my notes. I don't think the effects on students who are typically more of a struggle to get engaged would be any better.
After I realized AI was creating a trust problem between my students and me. I decided to take it a step further. At the start of this semester, I bought my students $1 notebooks from Amazon. Instead of submitting typed papers online, they are journaling by hand to analyze what they're reading.
...
The feedback was positive too. In anonymous comments from students evaluating the course, I've had numerous responses along the lines of: 'I actually hated that we weren't going to have laptops. But now I'm glad the professor actually did that. It was way more engaging not to have technology in the class.'
...
The switch to the notebooks has added about 2 hours to my workweek, but it's meant I'm having much more fun with my students. Two students asked if they could decorate the front page, and I've suggested we could even have an art contest at some point.
...
Before AI, a much smaller group of students was cheating the system by paying people to write their exams. The accessibility and affordability of generative AI mean that this problem has simply become more widespread.
A few years ago my middle schooler had to write a one page paper on something, and instead turned in 10+ pages of AI slop. Her punishment? She had to write the paper out by hand and turn it in,...
A few years ago my middle schooler had to write a one page paper on something, and instead turned in 10+ pages of AI slop. Her punishment? She had to write the paper out by hand and turn it in, and if it happened again she would only be allowed to turn in hand written assignments. She could still use the computer for research, drafts, etc, but the final copy had to be hand written. I thought that was a great middle ground, and a way to still get the benefits of technology.
That’s kind of funny because growing up one of my least favorite things about writing assignments was having to type them up neatly on my mom’s digital typewriter after I already hand-wrote and...
That’s kind of funny because growing up one of my least favorite things about writing assignments was having to type them up neatly on my mom’s digital typewriter after I already hand-wrote and proofread my draft.
And that was WAY better than having to do it on the even older mechanical typewriter because I had a delete key and could edit in line without having to commit each keystroke to print.
Yeah I'm not really opposed to students using technology, especially as a means of drafting or looking up info. Hell even if they use an LLM to generate a response and then hand-write that they're...
Yeah I'm not really opposed to students using technology, especially as a means of drafting or looking up info. Hell even if they use an LLM to generate a response and then hand-write that they're still physically writing and in some ways internalizing what they're copying.
How does she feel about it a few years removed from the event?
She still thinks it was 'totally unfair', but she also hasn't turned in any AI nonsense since. I think it was an appropriate response from the school, and at no point did they outright say 'no...
She still thinks it was 'totally unfair', but she also hasn't turned in any AI nonsense since. I think it was an appropriate response from the school, and at no point did they outright say 'no AI'. We will see how she goes as things get harder towards the end of high school.
Welp, this is my nightmare to be honest. I have dysgraphia and so writing by hand is messy, I make a lot of mistakes, and just an all around frustrating experience. I was lucky to get a dysgraphia...
Welp, this is my nightmare to be honest.
I have dysgraphia and so writing by hand is messy, I make a lot of mistakes, and just an all around frustrating experience. I was lucky to get a dysgraphia diagnosis, a LOT of kids don't. The only reason I did was because my dad was dyslexic so they got me tested for that early on.
I wasn't dyslexic at all, but I did have dysgraphia. And the interesting thing about Dysgraphia is that I have immense issues writing by hand legibly and clearly, but I can type absolutely flawlessly and extremely fast with no effort and retain knowledge that way. Oddly enough I can't draw a straight line or round circle to save my life, but I can design and do artwork on a computer perfectly fine and have made a career out of being a designer and artist. Brains are weird.
In middle school I even had this thing called an AlphaSmart that would let me type out my notes.
So I don't know, maybe we should look into more analog tech solutions too? Like offline word processors like the Alphasmart or purpose built laptops that only connect to an approved classroom network and doesn't have simple file sharing and won't work online outside the classroom. So your forced to turn in work physically in class with the laptop when the instructor allows a "turn in" window.
Actually that's a business idea in this day and age of AI. Creating a low-cost device for classrooms that doesn't have full online or general file sharing capabilities, so students are forced to use the device for notes and assignments, and has added convenience features in the classroom when connected to the closed classroom network such as interactive lessons, question asking/handraising, access to lesson materials, etc. 🤔
Edit: You could also have the devices have the ability to track keystrokes since it's not a general use device without internet connectivity, so no one should be using it for personal or private use. You could see if a paper a student wrote was "all-at-once" indicating they were just copying it from somewhere else like AI, or if there were natural pauses, re-writes, and editing happening.
You could write a software that tracks that and ads a score to the assignments they turn in, so if there's any question as to if it's AI or not, the file can be inspected to see if it was written all at once or not.
Edit 2: AND this kind of thing could be used in coding or computer classes too.
Also, to create redundancy you could design the software and hardware in a way that it's encrypted but also easy to switch out and recover in case a device breaks. Hell, you could even allow backups to happen but only in class. Meaning that in class with the device connected, the device gets backed up to the school's servers.
Edit 3: For a much more adoptable model that is easier to start up, but just a bit less secure given it's tracking keystrokes on personal devices even if it's only within the software, you could create word processor software intended to be standardized within the education industry, that tracks keystrokes and edit history within the software only.
So when turning in your papers, the file you turn in also includes the entire edit history of the paper as well. You could create a signing system to detect tampering, but basically it'll show the full history showing whether you copy and pasted the entire thing or wrote it by hand, or if you just sat there and typed everything out perfectly at once in one sitting, or if the paper was written and edited over the span of several days or weeks.
Software on the teacher's side can look at that history and determine if it looks suspicious.
Essentially using the way a person wrote the paper as the way to tell if AI was used instead of looking for arbitrary fingerprints within the writing itself.
Alphasmarts! Those things were (and still are) awesome. No internet. Simple LCD screen. Just for typing. Ten 'file slots.' As a novelist, I have used one (or a derivative of one) for years. Still...
Alphasmarts! Those things were (and still are) awesome. No internet. Simple LCD screen. Just for typing. Ten 'file slots.'
As a novelist, I have used one (or a derivative of one) for years. Still wonderful devices. Truly, sometimes less is more.
Something like a Remarkable tablet would work great, provided you could disable file hosting service integration or lock them to a university owned account. Something with a physical keyboard...
Something like a Remarkable tablet would work great, provided you could disable file hosting service integration or lock them to a university owned account. Something with a physical keyboard would be much better for students that struggle with handwriting though
Some sort of cyberdeck that's basically a typewriter would work well for this. I think tracking edit history is a bit of a garden path; better to stick with simple systems that inherently...
Some sort of cyberdeck that's basically a typewriter would work well for this. I think tracking edit history is a bit of a garden path; better to stick with simple systems that inherently discourage cheating instead of invading everyone's privacy or building some brittle Rube Goldberg machine.
While it's less exciting than a dedicated typing device, I think you can pretty much get the desired result with a managed chromebook in locked mode. If that's not sufficient, then it's just a...
While it's less exciting than a dedicated typing device, I think you can pretty much get the desired result with a managed chromebook in locked mode. If that's not sufficient, then it's just a software problem to solve.
I would like to think that a simpler dedicated word processor device could be cheaper, but the more I think about it the more I think in reality it's probably hard to beat the economies of scale of chromebooks. I think even a dumber device ultimately ends up having most of the components of a laptop anyway. Using a worse display, processor, battery, and storage doesn't reduce the cost that much, since you still need them, and chromebook components are already cheap. Although the existing modern Alphasmart clones are tailored towards a more premium design and audience, they're a lot more expensive than chromebooks. Other existing cyberdecks seem about as cheap as the cheapest chromebooks. My guess is the price of a system with all the needed functions probably can't be reduced to be significantly less than a chromebook.
If it were me putting the idea to paper, I'd sell the system to the school overall and then students can check the device out similar to textbooks, and turn it back in at the end of the semester...
If it were me putting the idea to paper, I'd sell the system to the school overall and then students can check the device out similar to textbooks, and turn it back in at the end of the semester and get a deposit back or something.
I'd design it in a way that the parts are very easily and inexpensively sourced and that the parts can be swapped out fairly easily by low level IT departments.
Students would be able to use them in multiple classes throughout the semester.
I mean, now that cyberdecks are being mentioned, I think I could build one of these with probably $100-$150 off the shelf with a 3D printed frame, so if I was mass producing them I could probably source everything for a lot cheaper than that wholesale. It probably wouldn't need more memory than a basic Raspberry Pi to be honest.
And if I'm selling them to be a reusable device for schools, I can make them a bit nicer for longevity.
And you wouldn't need to reinvent the wheel as far as software either. It'd basically be an office suite with under-the-hood user action history and a proprietary human scoring algorithm.
You'd need enough memory to do screensharing and some basic encoding/decoding a stream, but again a basic Raspberry Pi compute module can pull that off, so I'm sure there are other less open options that it'd work for.
I'm curious, what kind of privacy would need to be protected on a device only intended for classroom assignments with no open access to the internet? Like a device only made for writing and...
I'm curious, what kind of privacy would need to be protected on a device only intended for classroom assignments with no open access to the internet? Like a device only made for writing and turning in assignments?
Or on a piece of proprietary software that only tracks history within the app, but is also solely intended for classroom assignments. Although I get how the privacy conscious would be skeptical of installing something like that.
I'd also be curious if there was a way to write some kind of internal scoring system that is generated without storing or tracking actual user input history, but takes into account and builds a score based on the same things, and that score is what is attached to the file. Might be easier to game the system in that case, finding tricks and stuff.
I like the solutions you propose! I'm not sure how workable they are in terms of implementation both for me as an individual and across a university, but they seem fairly feasible in terms of how...
I like the solutions you propose! I'm not sure how workable they are in terms of implementation both for me as an individual and across a university, but they seem fairly feasible in terms of how they would work in the moment.
I totally get your concern about disabilities, and I would absolutely allow any student that have accommodations for technology take advantage of that.
Yeah haha I'm just kind of spitballing ideas as I enjoy doing that. There'd be some pitfalls to it, buuuut then again we've got billionaires selling AI, I think I could sell a college on a digital...
Yeah haha I'm just kind of spitballing ideas as I enjoy doing that. There'd be some pitfalls to it, buuuut then again we've got billionaires selling AI, I think I could sell a college on a digital solution that allows AI and education to somewhat coexist without AI threatening the integrity of the classroom.
I just wish I had the resources to do something with the ideas. haha
I have a couple of anecdotes about this approach. When I was in college years ago, the new shit was a calculator that could derive and integrate for you. I bought one thinking it was cool, but...
I have a couple of anecdotes about this approach.
When I was in college years ago, the new shit was a calculator that could derive and integrate for you. I bought one thinking it was cool, but then I failed calculus not once, but twice. During the second time I was slowly failing it, I ended up leaving it in a classroom on accident and never saw it again. The third time I aced the class because I didn't have anything to bail me out and had to actually learn it; never had a problem with it after that.
When I was working on my pilot's license, my old-timer instructor taught me to do everything by hand. Use a manual flight computer, paper charts, flight plan by hand, that kind of thing. I trained on a plane with steam gauges when the club had plenty of planes with fully digital avionics. The weakest part of my game was flight planning because there were so many digital aids that obfuscated the act of flight planning that I never really learned how to do it. Once I dropped those and went back to pen and paper, I sailed through my exam. My instructor had this motto: learn the hard way so you can graduate to the easy way.
I'm not a teacher or professor but I've been following discussions about the ai problem in the classroom. Some people claim to have substituted oral presentations for research papers. I think your...
I'm not a teacher or professor but I've been following discussions about the ai problem in the classroom. Some people claim to have substituted oral presentations for research papers.
I think your notebook idea is a fantastic idea. Personally I would have struggled because I have terrible handwriting but for most students I don't see much of a downside.
I actually do have an oral group presentation at the end of this semester students will do that works in tandem with a typed paper. The idea is that I can't really verify whether or not a student...
I actually do have an oral group presentation at the end of this semester students will do that works in tandem with a typed paper. The idea is that I can't really verify whether or not a student has used an LLM to write their paper, and if they do, having to talk about it in front of class will force them to at least read what they wrote.
I imagine that I will struggle a lot too because of terrible handwriting in terms of grading it, but that would be a burden I'd be willing to bear :)
As someone who works writing papers for people, I was scared that the AI revolution and physical paper approaches like the one described here would kill my work. Turns out it is about the same it...
As someone who works writing papers for people, I was scared that the AI revolution and physical paper approaches like the one described here would kill my work. Turns out it is about the same it has always been.
There are AI detection tools, so flat-out copy-pasting does not work, and between fact-checking the AI for hallucinations and rewriting the paper to make it sound human, it ends up being about the same effort as writing the paper yourself, especially for longer 3+ pages papers, so customers continue to hire me. In the case of hand-written papers, the customer simply "hand-prints" my paper with a good old pen.
Therefore, as a means to prevent cheating, this is a waste of the honest students time, and also a waste of valuable professor's time (reading handwriting is significantly slower than reading on a screen)
However, I will note that AI has elevated the quality of my work. You can do so much with it that is not "write my essay". You can ask it for alternate theses, to find obscure facts and data that are hard to find and easy to check, to provide feedback on your paper, to contrast your paper against the rubric, to find factual errors and omissions, to organize your brainstorm into structured paragraphs, and so much else.
If you are smart about it, you can use AI academically and increase the quality of your work while remaining 100% legit since you are not plagiarizing the LLM's output. It's like having a personal teacher. If you are dumb about it, you will copy-paste or replace words instead of properly paraphrasing AI and get found out.
To the professor's credit, studies do show that handwriting can improve retention by forcing deeper processing and involving the motor cortex and other parts of the brain not stimulated by plain typing, so it might genuinely help some students. But mandating it for everyone, to me, feels like punishing the majority to catch a few, especially when smart AI use is already legit and undetectable if done thoughtfully.
I guess the TL;DR of what I am saying would be:
A) Handwriting is ineffective to stop LLM or traditional cheating.
B) It increases workload for everyone for very minor retention benefits.
C) AI can be used responsibly.
So a better policy would be teaching students how to use AI well, rather than being a Luddite about it.
Given the rate of false positives, I can't ethically use AI-detection tools as a proof of a student cheating. Ultimately, I'm not concerned with students finding a prompt and copying that by hand....
Given the rate of false positives, I can't ethically use AI-detection tools as a proof of a student cheating.
Ultimately, I'm not concerned with students finding a prompt and copying that by hand. This has always been a problem and there is no clear solution for it. But, as you point out, handwriting seems to be the best way of learning and retaining information. If a student is forced to hand write an output from an LLM or someone they hired, they are still processing that information in some way as opposed to copy and pasting or do quick edits to a file (which many students do).
My interest in this, as was illustrated by Ayşe Baltacıoğlu-Brammer, is that I want to improve student engagement with the course. This isn't, and there is no, silver bullet solution to that. Some students will refuse and will find any way to work around it. But if I can have a net improvement with engagement and students take away new studying skills, spend less time on a computer, and think more deeply about the material then I would view that as a success.
As per your last sentence - I already can't cover all the concepts that I think is relevant to the course material. I'm not too interested in carving out more information to dedicate towards responsible and ethical AI use, and on a personal note, I think spending less time on computer screens is probably better for us so I'm OK being a bit of a luddite about this :)
One of the things I took from the article was that it creates a far higher barrier to entry to student's replacing study/learning with the use of AI or hired paper writers (where the student...
One of the things I took from the article was that it creates a far higher barrier to entry to student's replacing study/learning with the use of AI or hired paper writers (where the student trying to read and understand the output; removing themself from the loop). It is a LOT harder to procure a paper and then copy it out than it is to print off a procured paper, and I suspect that would change whether it is "worth it" to a lot of could go either way students.
I think this is the bigger discussion that needs to happen. Like it or not AI is here to stay, and we do need to start showing how to use it responsibly. The calls to flat out ban it just remind...
C) AI can be used responsibly.
So a better policy would be teaching students how to use AI well, rather than being a Luddite about it.
I think this is the bigger discussion that needs to happen. Like it or not AI is here to stay, and we do need to start showing how to use it responsibly. The calls to flat out ban it just remind me of the people from my childhood who wouldn't let their kids use computers or screens at all. Like sure, less screen time is good and I get where you are coming from, but a giant portion of the working / real world is computers and screens. It doesn't actually help to hide those things away. I think AI is in that same boat.
I'm not really opposed to the use of LLMs and other tools like it - I acknowledge that they can be very useful and I talk about some uses I am OK with in my course (e.g. a spring board for...
I'm not really opposed to the use of LLMs and other tools like it - I acknowledge that they can be very useful and I talk about some uses I am OK with in my course (e.g. a spring board for identifying drug targets to research for a condition, generating names for a drug). However, I also think it's important to demonstrate that you don't need to use these tools and that proper usage of the tools requires a foundation of certain skills/knowledge. We're seeing negative trade-offs in cognitive skills as people offload their thinking to these models, and so I don't think that forcing students to fall back on and develop those skills is going to harm them.
This kind of approach is likely to work well in the short term, particularly for students who over rely on AI to write basic prose for them. However, AI assistance is already opening a gap between...
This kind of approach is likely to work well in the short term, particularly for students who over rely on AI to write basic prose for them. However, AI assistance is already opening a gap between "cheaters" and skilled researchers. Smart people need to learn how to critique and marshal AI outputs. If teachers forsake modern tools, their students will be unprepared for the ways those tools will both enable and betray them. The floor is raising; it will become a mistake to also lower the ceiling.
These models have gotten way better in a short amount of time and there are many use cases for them to automate tasks and provide concise summary of material, there's no doubt about that. This may...
These models have gotten way better in a short amount of time and there are many use cases for them to automate tasks and provide concise summary of material, there's no doubt about that. This may be a lack of imagination, but beyond something like generating a quiz to test your understanding or using it as a smart research tool, I struggle to see how incorporating it into my lessons is going to facilitate the learning process as it pertains to developing their reasoning, which is what I'm interested in. I already tell them to use it as a starting point for identifying drug targets, but that it should be taken with a dose of skepticism because they can't always evaluate the correctness of outputs. I am looking out for workshops to provide some ideas on how to better incorporate LLMs into my class and assessments - I am not completely unwilling to use the technology, but I don't think I should be teaching students how to use it responsibly if I don't have a clear plan for it.
I lack any applicable knowledge in teaching, so this is likely not a useful direction, but may I ask if debating/correcting an LLM could be useful, if your intent is to develop students' critical...
I lack any applicable knowledge in teaching, so this is likely not a useful direction, but may I ask if debating/correcting an LLM could be useful, if your intent is to develop students' critical thinking skills? Not sure how well this could be pulled off, but for example:
In a mathematics-focused course, have the student "tutor" the LLM, which has been instructed to have made a mistake in its reasoning for an example problem, which led to an incorrect result. The student needs to understand what the LLM did wrong, and then offer the correct strategy for
In a philosophy-focused course, have the LLM take a poorly defensible position based on loose rhetorical grounds. The student should be able to identify the flaws in the LLM's arguments based on the "debate".
For my part, I'm mostly using them to help with learning concepts from recorded lectures in MOOCs: the communities for those have long since died off, and in university, I really leaned on being able to ask for clarifications at the end of each lecture. LLMs have been very helpful in that regard, especially if they cite their sources in their responses (which, as you note, are sometimes a wee bit ... creative).
I do like the idea of turning the student into a tutor, that's a creative idea for a study tool. I hadn't thought about setting a parameter of an LLM of being intentionally wrong - further proof...
I do like the idea of turning the student into a tutor, that's a creative idea for a study tool. I hadn't thought about setting a parameter of an LLM of being intentionally wrong - further proof there is a lot I have to learn about them. I was viewing it from the lens of, "They don't yet know the material, so they can't evaluate if the output is wrong". I'll try playing around with ChatGPT and see how I feel about it, thanks for sharing your ideas!
I hope it works out! If it helps, I've found that Claude and Gemini (to a lesser extent) do a better job of responding factually if you're able to upload content for them to pull references out of...
I hope it works out! If it helps, I've found that Claude and Gemini (to a lesser extent) do a better job of responding factually if you're able to upload content for them to pull references out of (and then explicitly tell them to cite from it). I've still had it hallucinate a lot, though, and you really need to urge them to "do research" in order to avoid frustratingly incorrect conclusions ...
(Gemini's current model is extremely bad about this, and gets mouthy if it's corrected ...)
Food for thought: I like this idea of banning technology in the classroom (excluding accommodations) and putting students on a track of hand-writing assignments, so much so that I'm considering implementing this for my 400-level course next semester. This is going to add a lot of work to my plate, but I think removing ways for students be distracted in class, no matter how good intentioned they are, and having to actually write out their thoughts (or even copying it down from an LLM) should hopefully improve engagement with the course.
Outside of disability and English-second language individuals, what are some potentials pitfalls of this approach? Would you have liked this as a student, if you were one that used your computer to take notes?
I type literally 10x faster than I can write, and half of my professors could not read my handwriting in early college. Im sure part of that is being out of practice, but still, you had been my professor and had done that, I probably would have bought a typewriter or taken my laptops wifi chip out of my laptop in front of you. I dont use AI, hope to never have to, but a computer is the only way I could have ever hoped to keep up with my notes. In some classes with fast speaking professors, even that wasnt enough, and I bought a dedicated voice recorder for those classes.
I certainly understand that through-put is much higher with typing for most people. The organization of the keyboard and the nature of typing is pretty much always going to beat out handwriting.
But when it comes to learning, going fast is not really all it's cracked up to be. It's generally better for us to slow down the learning experience because it helps encode the information more efficiently. A recent meta-analysis backs this up - handwriting notes tends to improve retention and leads to better academic outcomes.
For what it is worth, I hate this line of logic and it's caused me no end of friction. People learn in different ways, I did plenty fine typing out my stream of consciousness notes and despised the classes that forced me to "slow down" because their definition of slow was much much slower than mine.
I understand the good intentions here, but I would outright refuse to take a class that wouldn't let me type in this day and age. I also think it's doing a disservice to the students to not prepare them to use the tools they'll actually be using in real environments.
This feels like the kinda of academic optimization that's far better in ideals than reality.
That's fair, and I would be upfront about my policy at the start of class so students can drop if they want to. Ultimately, because people learn in different ways, no matter the approach and pedagogical techniques I use, some students will benefit while others will falter. Such is the reality of not doing individualized instruction. So I'm going to go where the evidence leads me on best practices to help the most amount of students.
I don't necessarily disagree that we should prepare students to use the tools that they'll be using in a job, but I view that as being subordinate to developing the foundational skills that will maximize the utility of those tools. And in my experience as an instructor, many students desperately need to practice working through their thought process and developing arguments. On top of that, I don't think I'm doing students a disservice if I don't let them type notes in my class - they already have plenty of experience with using a computer.
That's fine, but I don't agree your solution will actually lead to that, and it seems more like an arbitrary hurdle that's got research from some ideal setting rather than well tested.
I had a teacher in high school who said we should have to write only on unlined paper because that's what would be expected of us going into college. That was NEVER true, never helped, and made me drop out of the class because it unfairly punished people like me with bad handwriting as I'd get docked points for not following where the lines should have been (we were supposed to put lined paper underneath the white sheet to adhere).
She was a good teacher, and instead I chose to drop because of the arbitrary restriction. I don't think yours is objectively as bad, but I doubt you're going to see the results you want. I wish you luck with it though.
That certainly is a pretty annoying and harrowing rule to have. I wonder if it may have colored your perception of having to write things by hand? It certainly would make me aversive to the idea.
Hopefully it does goes well for my students. I have no problem going back on it if causes more issues than it's worth. Thanks for the discussion.
Its not that typing is faster, its that my writing is slow enough, and professors fast enough, that handwriting in early college often led to me only encoding 2/3rds of the information at all, leading me to doing horribly on tests if I wasnt able to use recording devices or borrow somone elses notes.
Just out of curiosity - did you do any reading of the material before class, or was the lecture your first exposure to it?
In my degree only half of the classes had textbooks, and only half of those that did ever told us what the material we should read was ahead of time. When provided, I much preferred reading the textbooks to the lectures and would read them both before and after, yes.
OK thanks for answering.
I don't require a textbook, but I do suggest one and I emphasize to my students that they should be reading about the material before coming to class that way when they hear new terms they aren't completely lost. I really don't want to make textbooks mandatory because of how expensive they are, but I think it's very beneficial to read about these concepts (barring some sort of reading/learning/visual disability).
Yup, textbooks make everything easier, but I hear you WRT the expense. I had been sailing the high seas for quite a bit before college, and during college I kinda became the local 'pdf dealer' in my program. I know over 50 people who got all of their textbooks from me.
To be fair, that's probably more of a professor issue than a you issue. Some professors seem hell bent on covering too much imo. It feels counter intuitive, but in my experience it's more effective to cover less in a lecture, take it slow, and provide additional out of class resources (and definitely not just out of class readings that everyone obviously actually reads) to fill in gaps.
Perhaps typing it first, at least in outline form, and then transcribing it to handwriting would allow for better retention.
Yeah ... I don't know how much I trust that meta-analysis.
For context: I'm used to seeing meta-analyses use a robust framework for rejecting studies that are poorly designed, or are clearly biased. The fact that they also included non-peer reviewed studies is kind of maddening, too. For example, why were remote courses removed from the analysis? They were fine with pre-recorded lectures in a lab, why not at home? The authors even note that inclusion bias could be screwing up several other meta-analyses which have shown no result, instead of demonstrating a superior medium (see
Findings from Previous Meta‑analyses).I don't have the brain space atm to read through this paper (How to interpret discrepancies in empirical results from educational intervention studies.) (pre-print here), but IMO at a skim it confirms my biases at least: that educational intervention studies are extremely difficult to design, they get blown out of proportion by people with an agenda, and that we basically know nothing and should stop taking away tools from students.
(edit) Sorry, just clarifying my point, because I'm pretty scatter brained atm. People have been complaining about "kids these days" since time immemorial. Equally, seemingly everyone has an opinion about how kids should be raised, or what is the morally correct way of doing things. I think it's all bupkis and pseudoscience, and that you'd find a dramatically larger effect size on student performance with other, less contentious interventions: pay teachers more, decrease class sizes, change incentives towards learning instead of rote memorization/cheating.
(and for context, I can handwrite ridiculously fast anyways, and preferred switching between longhand and typed notes in university, based on the class. Calculus, discrete math, etc. got the wacom tablet, whereas psychology, linguistics, etc. got the keyboard. Depriving students of tools only makes them less capable humans overall)
I'm not too interested in litigating all of the research on this because I'm not that committed to the idea that things must be handwritten in order to be learned. I know many people encode information from typing and by reviewing the notes they typed. There is no one-size fits all approach to note-taking and education that could be implemented in a classroom. The main reason I subscribe to the idea of writing things down as a good technique for students is because my pet hypothesis is that forcing students to slow down is better for retention (which I seem to recall is backed by some study, but I don't have the time to dig for it at the moment), and I think that most students aren't taking that time to slow down. There are many perverse incentives for why they don't slow down, such as overloading their schedules with too many classes and not having enough time to dedicate to studying while balancing personal life and work.
Regarding your last point about less contentious interventions - I don't think the first two are less contentious, otherwise we would simply do those, no? Yes, I would love to have less students so that I can work with them more directly and paying teachers more would hopefully increase retention of good teachers and encourage more people to join the profession, but we pretty clearly don't want to do that as a society right now and so I need to think about other ways of trying to improve my pedagogy given the system I'm in.
No worries. I'll bow out; I was concerned that people in this thread were reading a study, thinking it reflected reality, and were attempting to act in a research-driven manner. If we're all kinda winging it, that's fine. Rigor isn't necessary anywhere, it's just annoying to see people wag scientific papers around to prove their point, when vibes were always a perfectly reasonable way to ground your decisions.
I won't dig into your perspective; I disagree with it, but I'm doing so based on different grounds than you (i.e. it seems that the field can't agree on anything, so claims can't be made either way), so it's not terribly productive to discuss it further.
Yep, fair enough. I wasn't trying to suggest that you do anything different. Mostly, I'm suggesting that Business Insider stops posting articles which shape a narrative that actually laptops are the problem w.r.t. academic achievement. Not systemic issues in the school system.
(life issues are draining my brain at the moment, and I apologize for any bluntness in the above comment. I'd normally give it a few more editing passes, but I have several issues to attend to. Good luck with your class; I'm glad you're trying to do what's best for your students)
Truly, thanks for sharing your thoughts. I do agree that people will drop papers all the time as justification without much further thought - the main reason I wanted to share the meta analysis was that it's not a crazy idea to encourage students to shift to handwriting (but as you pointed out, the evidence really is murky, thanks for sharing that article in your previous response. I intend on reading it over spring break). I just didn't want to get into an intense discussion about what constitutes valid evidence for a meta-analysis because that can be quite a quagmire.
Certainly Business Insider, and other publications, will be looking for simple explanations for complex issues. These ultimately are caused by systemic issues, and I am totally sympathetic with you on that. If I was the education tzar and had a lot of power to change the education system, it would look drastically different than what it is right now. One of those things would not be the tzar and actually having a panel of informed people to help make decisions.
Handwritten notes being shown to improve outcomes on a population level does not entail that it's an improvement in every individual circumstance. You can't wave around results like that as though they universally apply to individuals. That's not how science works. And I say this as someone for whom handwriting notes does improve things, personally.
I didn't claim that it universally applies - I'm aware there is no one size fits all approach to learning via note-taking as I've mentioned in other comments on this post. And in fact, I acknowledge that in the comment you're replying to by explicitly typing out "generally" and "tends to". However, it seems from the data I've read that more people benefit from handwriting notes than other methods.
In a similar vein, not everyone sees benefits from exercise when it comes to the numerous health outcomes we associate with it. However, I don't think it's faulty to suggest that generally exercising is a good idea.
I don't think that citing that meta-analysis in response to someone saying they struggle with handwriting notes and do better when they can type amounts to just observing trends, but is instead giving advice to an individual who has already expressed their lived experience about how these methods affect their note-taking. Even if the research on this were rock-solid (and I'm not convinced it is), I think it may be worth imagining how frustrating a response that is so dismissive of your own experience with notetaking and your own brain would be, even if they pay lip-service to there being individual differences. I have friends with chronic illness who have the same experience with people constantly insisting they should try yoga even after they've mentioned that yoga would actually exacerbate their particular disabilities, so perhaps exercise isn't the worst comparison. It's tone-deaf, imo.
As other comment in this thread noted, I think this is a feature not a bug.
But I do understand my frustration : my writing style is messy and iterative ; there is a cost in writing stuff down over and over again.
Isn't that the solution? I'm not sure there's much value in being a stenographer. Notes should be more condensed than natural human speech, which is not particularly information dense by default.
Of course, professors can refuse to allow recorders, but they can refuse to allow anything - I once had a professor refuse to allow us to write anything down.
Personally, I am totally fine with student's voice recording lectures. I also agree with what you're saying about notes - you shouldn't be handwriting every single thing that I'm saying or that appears on my lecture slides. Take out main points and use those as a foundation for understanding the other material.
That really sucks that your professor wouldn't allow you to write anything - what was their rationale?
Taking notes inherently implies a structure where there's a speaker and listeners. That professor wanted a more communal form of teaching.
It was a pretty niche class - a class on gnosticism and other early esoteric christians sects. There were like 8 students, so we just sat in a circle and it was expected that if we didn't understand anything immediately, rather than try to take notes, we would simply ask for clarification.
I didn't mind, really. Was a good experience, if anything.
Ah OK - yeah I can definitely see that working. There wasn't an explicit rule about that in my graduate level insect physiology course, but the instructor set it up in a way similar to what you're describing where taking notes didn't really make sense. I think that works really well for small class sizes.
Isn't it a problem only when writing absolutely everything ? Like doing a transcript ?
I can only speaks from my point of view, but I never encountered a class where I needed to write everything.
Whenever I handwrote something for my studies, it was always by summarize and rephrase what I've listen to, and choosing actively what was important to take note of, and what was optional to note.
And I don't remember one time re-reading my notes in details. It was a tool to help my brain re-connect information together.
The act alone of handwriting became a good way to remember the information I've just learned for way longer.
I get where you are coming from, but I don't think anyone is worried about ai for note taking, at least not from a cheating angle :)
My only issue with AI note-taking is that physically handwriting the notes is better for retention - beyond that I would not care at all if a student was using a voice-to-text dictation as long as it helps them better understand the material.
I was a TA for awhile, and handwriting is a huge barrier to being able to read and evaluate a student's work. I think there are definitely advantages to doing assignments by hand in class - it forces students to complete assignments there and to potentially discuss course content with peers - but it was so miserable on the grading front. I would not do it if you have more than ~30 students at the most, unless you're cool with tormenting yourself and/or your TAs!
I had a few Francophone students at my Anglophone university, and I do think they benefitted from being able to practice writing by hand, especially if they had time to work with peers - having handwritten assignments slows everybody down, and having a more group-focused context can allow them to discuss ideas and support each other. So having ESL students isn't necessarily going to be a problem, depending on how the course is structured and how much they can work with other students. If they are all working independently, that may add additional barriers, but it might be worth chatting with those students to get their perspectives.
The class I'm planning on implementing this in is ~45 students. I can't keep up with grading a weekly assignment, so I'm thinking of maybe having them submit something every 4 weeks around exam time. Hopefully that can give me a quick turn-around in terms of grading and getting it back to them without them having to rely on it for studying.
You can always test it out and see how it goes! I think it's great that you're thinking about how to achieve good learning outcomes for your students, and I think trial and error can be helpful to refine your vision. Good luck!
Doctors are notorious for having bad handwriting and have been for a really long time. I think legibility has more to do with how we’re taught to write.
My problem here is that we shouldn't be optimizing our education system for the top 10% of performers if it means worse outcomes for the bottom 50%. Having a capable, educated workforce benefits everyone.
The compromise I see here is an extension of something we've already done for a long time: putting high performing students in accelerated education programs. A student who demonstrates above average proficiency and self-discipline could be placed in classes where assistive technology is more readily available.
I agree with you on this - we need to try and reach and teach the majority of people. Ultimately, we need more individualized instruction plans based on where the students are at. I don't think we can do it truly on an individual-by-individual basis (although that would be ideal), but perhaps having several streams of course concepts/practice extending and applying the idea of AP classes in high school - at all levels of education.
Even when I was in practice my handwriting was so bad and so slow compared to my typing speed. It was such a breath of fresh air to finally swap over to typing out papers and such. It sucks so much now that its becoming more of a liability to teachers, but I can't imagine being forced to going back to hand writing out a 10 page paper, I'd cry.
It might work, depending on the type of class and the subject matter.
However, I think you would need to make some accommodations for the change in approach. I agree with the comments about handwriting being slower and more difficult to read.
Some ideas:
If you are only focused on note taking, since some people will be slower than others you can adjust the content so there is less to write, or provide some type of handout so they have some notes already.
If it is an assignment, I don't think techniques like having a minimum word count would work. A requirement that is inherently easier to do on a computer can end up distracting from the objective of the assignment when asked on a handwritten assignment.
If you are going to read their handwriting, it would be good to increase the chances that their handwriting is as neat as they can make it. For example, some people will write neater when they have more time to write. Also, consider an "intro to note taking" session that can include topics like useful shorthand so ppl dnt thnk thy hv 2 fully spell out evry wrd.
Thanks for sharing these ideas! I like the idea of doing a short crash-course on note-taking in my introduction/go over the syllabus class.
I certainly would not have a minimum word count for written assignments. I would probably focus on having some really clear rubrics that highlight what all they have to do in terms of addressing questions or engaging in discussion. Fortunately, I have all summer to think about and implement these changes.
Off the cuff, I would speculate this works well in a lot of classes but would potentially fall completely flat with a lot of CS related courses. Anything where the topic directly involves a computer or electronic device. There are probably variants of the notebook-only approach that would work for that as well, but might be more difficult to figure out and for the students to adjust to.
I never took university-level classes, but I think I would probably have hated it (like the student comments from the article) but ultimately been grateful that it was done that way.
You can absolutely do hand-written exams and assignments for most CS classes, and many of mine did, especially in the beginning of the degree. I think it was a good choice and helpful in the long run. A bit tedious? Sure, but it really forces you to learn programming logic and ingrain a deeper understanding of the topics in your mind. A lot of classes (like data structures) don't necessarily need actual programming right away either.
My favorite CS professor started off his intro class with extolling the virtues of hand writing notes, and he was absolutely right. The few of us that took this to heart ended up doing very well throughout the CS program, and generally seemed to be above average in our understanding (though tbf there always seemed to be a pretty bimodal split this way, especially when it came to the weed out classes). In addition, using a laptop during class was just far too distracting for me, so it was extremely beneficial in regards to forcing me to pay attention (or at least more attention than I would've).
Every single CS class I had used handwritten exams, and they were weighted to be the majority of the grade.
At the freshman level, this often involved "here's a printed Javadoc" or "I hope you memorized C functions" and having to write out a program. In higher level classes, which quickly become about theory and not writing code, you end up doing things like labelling states of registers on a toy CPU based on assembly instructions, or drawing diagrams of how a red-black tree is changed at each step of an algorithm.
Honestly, programming is a prereq to start a CS degree. It's not what the major is about, and if you're not already decent at it, the limited classes that focus on it aren't going to do much to remedy that.
There are certainly environments where you need a computer or some sort of device - a programming course for instance would require it and I would not expect students to do otherwise. But, maybe in the recitation where you are talking about programming concepts you wouldn't necessarily need to have a computer.
I was a CS major and the classes which required a computer typically took place in the lab, where there were desktop computers available for everyone.
Towards the end of my degree (2018) there were maybe 1-2 times that we ended up in a lab without computers because of some kinda scheduling mistake.
Its actually kinda weird to me that students these days almost always have their personal laptops in the lecture hall. Thinking about it now, I have tons of hand written notes from my CS major, I didn’t use my laptop during lectures often. I relied on my laptop mostly for homework, because although I could use the lab computers for my assignments, I liked my programming setup better and doing homework in my apartment beats having to stay on campus.
When I grew up there were no electronics in class. The best students could feel some level of exhaustion in the muscles of their hand and wrists. It wasn't anything serious as far as I know. I had awful handwriting and sometimes couldn't read what I wrote. So lookout for those kids, teach them how to write correctly. And get adequate stationery for the left-handed like me. Other than that I think it's a great move.
I'm really thinking about high school and below here. Film school didn't require much note taking from me. I'm not sure how that would work for other majors.
Even as a student who does perform better when I limit myself to handwriting my notes and don't bring my laptop to class as a potential distraction, I don't think wholly banning them is acceptable. Firstly, you cannot ignore disability and ESL as these will be hugely salient in almost any class you're in. Secondly, even people without a disability may have valid reasons they prefer to take notes on a digital device -- much like how I use subtitles despite not being hard of hearing. Just because they're a disability aid doesn't mean they can't often be helpful to people who aren't disabled too.
Thirdly, frankly, in a 400-course it feels incredibly patronizing and like being treated like a child. If it's a seminar and is more discussion focused anyway, I can see a no-laptops policy making sense (although even then, that's disruptive to those who save the readings as digital pdfs rather than printing them out, which was common in my field and has similar concerns when it comes to disability), but for a lecture or similar course where you're expected to take notes, it comes across as treating students like little children who are only allowed to do what you trust them to do. I remember when I first attended a college course as part of post-secondary enrollment, I was in awe of the fact that I was allowed to pull my laptop out when needed without the teachers acting like cops about it, because that was all I knew from the strict rules in high school. Being treated like an adult who could make their own choices about how they learned made me respect my professors more, and I eventually came to my own conclusions about what note-taking methods worked best for me. If I had been forced to take "analogue" notes by my professors, I would have resented it and never learned the ways that they work better for me personally, and/or would have avoided taking notes at all out of spite. And I'm a student whose disability means I benefit from taking handwritten notes. Imagine how much worse this feeling of overbearing patronization would be when compounded by a disability that makes it harder with handwritten notes.
Regarding your first point - I never claimed that I would prevent students that have a disability or ESL from using these devices? I'm aware that their learning may be greatly impaired if they're not able to use electronic devices. Me phrasing my sentence that way is to acknowledge that there would be no question about them bringing it in.
That is a fair point regarding the rest of what you wrote. I imagine that if I do implement this rule some students would build up some resentment. Some students also, as noted in this article, might develop an appreciation and get more out of the class. No matter the expectations and interventions in the classroom, some students will benefit from a given pedagogical technique while others might suffer from it.
I think in principle you want to accommodate those who are disabled or have other barriers like ESL involved, but your descriptions in other comments of how you'd plan to present and implement this policy indicate that this would not be free of negative effects on some of these people in practice. How do these people get accomodation so that they're exceptions to this policy? Do they need to go through your university's accomodations office? Do they need to personally ask you and reveal potentially private information about their disability? Even if these things are not the case and you grant accommodations to anyone no questions asked, these are barriers that accomodations typically require and the prospect of them existing here will probably discourage some subset of students who need them from asking for accomodations at all -- they'll either suffer through the course as best they can and perform worse than they otherwise could, or they'll drop your course and lose out of the opportunity to become engaged and learn what you're teaching in the first place. This is going to be especially true for those who don't have a formal diagnosis, as they'll have more experience being denied accommodations (or less experience fighting for them) than other students. ESL students are in a different circumstance but they have similar pressures to not ask for accommodations when a professor places a rule like this in the syllabus, even if accommodations are available. The existence of the policy itself will discourage people from asking for "exceptions" as accommodations. You simply cannot make a blanket policy like this without collateral effects on some of the people. If you choose to implement this policy, you need to be well aware that you will be denying access to some disabled people, even if that's not your intention. There is no version of this policy that won't affect some of those people, and implementing this policy entails you believe its benefits are worth that.
It's true that different pedagogical styles help and harm different students differently. But I don't think it's remotely clear that the benefits of this one for some students outweigh the potential harms for others. And even if they're roughly equal, why would you choose the policy that demonstrates a paternalistic lack of faith in your own students rather than one that gives them even a small amount of freedom to discover for themselves what works best for them when it comes to how to best study and take notes? I would never have learned that I personally do better taking handwritten notes without doing different things for different courses (sometimes on different days in the same course) and discovering this through experience. I, like your students, am an adult who can make my own choices even if they result in me not maximizing how much information I retain, and the freedom to make my own choices rather than being forced to take notes in the style my professor believes is better based on some meta-analysis they read is part of being respected as an adult who can learn and engage with the material rather than as a naughty child who must be reigned in so that they aren't disruptive in class. I would chafe at that bit and I don't believe that the positive effects of handwritten notes would remotely equal the negative effects of me resenting you in terms of my engagement with your course and the material. And frankly I'm probably the best case scenario, as I'm a generally highly engaged student who does benefit from handwriting my notes. I don't think the effects on students who are typically more of a struggle to get engaged would be any better.
Some key snippets from the article:
...
...
...
A few years ago my middle schooler had to write a one page paper on something, and instead turned in 10+ pages of AI slop. Her punishment? She had to write the paper out by hand and turn it in, and if it happened again she would only be allowed to turn in hand written assignments. She could still use the computer for research, drafts, etc, but the final copy had to be hand written. I thought that was a great middle ground, and a way to still get the benefits of technology.
She was less impressed.
That’s kind of funny because growing up one of my least favorite things about writing assignments was having to type them up neatly on my mom’s digital typewriter after I already hand-wrote and proofread my draft.
And that was WAY better than having to do it on the even older mechanical typewriter because I had a delete key and could edit in line without having to commit each keystroke to print.
Yeah I'm not really opposed to students using technology, especially as a means of drafting or looking up info. Hell even if they use an LLM to generate a response and then hand-write that they're still physically writing and in some ways internalizing what they're copying.
How does she feel about it a few years removed from the event?
She still thinks it was 'totally unfair', but she also hasn't turned in any AI nonsense since. I think it was an appropriate response from the school, and at no point did they outright say 'no AI'. We will see how she goes as things get harder towards the end of high school.
Welp, this is my nightmare to be honest.
I have dysgraphia and so writing by hand is messy, I make a lot of mistakes, and just an all around frustrating experience. I was lucky to get a dysgraphia diagnosis, a LOT of kids don't. The only reason I did was because my dad was dyslexic so they got me tested for that early on.
I wasn't dyslexic at all, but I did have dysgraphia. And the interesting thing about Dysgraphia is that I have immense issues writing by hand legibly and clearly, but I can type absolutely flawlessly and extremely fast with no effort and retain knowledge that way. Oddly enough I can't draw a straight line or round circle to save my life, but I can design and do artwork on a computer perfectly fine and have made a career out of being a designer and artist. Brains are weird.
In middle school I even had this thing called an AlphaSmart that would let me type out my notes.
So I don't know, maybe we should look into more analog tech solutions too? Like offline word processors like the Alphasmart or purpose built laptops that only connect to an approved classroom network and doesn't have simple file sharing and won't work online outside the classroom. So your forced to turn in work physically in class with the laptop when the instructor allows a "turn in" window.
Actually that's a business idea in this day and age of AI. Creating a low-cost device for classrooms that doesn't have full online or general file sharing capabilities, so students are forced to use the device for notes and assignments, and has added convenience features in the classroom when connected to the closed classroom network such as interactive lessons, question asking/handraising, access to lesson materials, etc. 🤔
Edit: You could also have the devices have the ability to track keystrokes since it's not a general use device without internet connectivity, so no one should be using it for personal or private use. You could see if a paper a student wrote was "all-at-once" indicating they were just copying it from somewhere else like AI, or if there were natural pauses, re-writes, and editing happening.
You could write a software that tracks that and ads a score to the assignments they turn in, so if there's any question as to if it's AI or not, the file can be inspected to see if it was written all at once or not.
Edit 2: AND this kind of thing could be used in coding or computer classes too.
Also, to create redundancy you could design the software and hardware in a way that it's encrypted but also easy to switch out and recover in case a device breaks. Hell, you could even allow backups to happen but only in class. Meaning that in class with the device connected, the device gets backed up to the school's servers.
Edit 3: For a much more adoptable model that is easier to start up, but just a bit less secure given it's tracking keystrokes on personal devices even if it's only within the software, you could create word processor software intended to be standardized within the education industry, that tracks keystrokes and edit history within the software only.
So when turning in your papers, the file you turn in also includes the entire edit history of the paper as well. You could create a signing system to detect tampering, but basically it'll show the full history showing whether you copy and pasted the entire thing or wrote it by hand, or if you just sat there and typed everything out perfectly at once in one sitting, or if the paper was written and edited over the span of several days or weeks.
Software on the teacher's side can look at that history and determine if it looks suspicious.
Essentially using the way a person wrote the paper as the way to tell if AI was used instead of looking for arbitrary fingerprints within the writing itself.
Alphasmarts! Those things were (and still are) awesome. No internet. Simple LCD screen. Just for typing. Ten 'file slots.'
As a novelist, I have used one (or a derivative of one) for years. Still wonderful devices. Truly, sometimes less is more.
Something like a Remarkable tablet would work great, provided you could disable file hosting service integration or lock them to a university owned account. Something with a physical keyboard would be much better for students that struggle with handwriting though
Some sort of cyberdeck that's basically a typewriter would work well for this. I think tracking edit history is a bit of a garden path; better to stick with simple systems that inherently discourage cheating instead of invading everyone's privacy or building some brittle Rube Goldberg machine.
While it's less exciting than a dedicated typing device, I think you can pretty much get the desired result with a managed chromebook in locked mode. If that's not sufficient, then it's just a software problem to solve.
I would like to think that a simpler dedicated word processor device could be cheaper, but the more I think about it the more I think in reality it's probably hard to beat the economies of scale of chromebooks. I think even a dumber device ultimately ends up having most of the components of a laptop anyway. Using a worse display, processor, battery, and storage doesn't reduce the cost that much, since you still need them, and chromebook components are already cheap. Although the existing modern Alphasmart clones are tailored towards a more premium design and audience, they're a lot more expensive than chromebooks. Other existing cyberdecks seem about as cheap as the cheapest chromebooks. My guess is the price of a system with all the needed functions probably can't be reduced to be significantly less than a chromebook.
If it were me putting the idea to paper, I'd sell the system to the school overall and then students can check the device out similar to textbooks, and turn it back in at the end of the semester and get a deposit back or something.
I'd design it in a way that the parts are very easily and inexpensively sourced and that the parts can be swapped out fairly easily by low level IT departments.
Students would be able to use them in multiple classes throughout the semester.
I mean, now that cyberdecks are being mentioned, I think I could build one of these with probably $100-$150 off the shelf with a 3D printed frame, so if I was mass producing them I could probably source everything for a lot cheaper than that wholesale. It probably wouldn't need more memory than a basic Raspberry Pi to be honest.
And if I'm selling them to be a reusable device for schools, I can make them a bit nicer for longevity.
And you wouldn't need to reinvent the wheel as far as software either. It'd basically be an office suite with under-the-hood user action history and a proprietary human scoring algorithm.
You'd need enough memory to do screensharing and some basic encoding/decoding a stream, but again a basic Raspberry Pi compute module can pull that off, so I'm sure there are other less open options that it'd work for.
I'm curious, what kind of privacy would need to be protected on a device only intended for classroom assignments with no open access to the internet? Like a device only made for writing and turning in assignments?
Or on a piece of proprietary software that only tracks history within the app, but is also solely intended for classroom assignments. Although I get how the privacy conscious would be skeptical of installing something like that.
I'd also be curious if there was a way to write some kind of internal scoring system that is generated without storing or tracking actual user input history, but takes into account and builds a score based on the same things, and that score is what is attached to the file. Might be easier to game the system in that case, finding tricks and stuff.
I like the solutions you propose! I'm not sure how workable they are in terms of implementation both for me as an individual and across a university, but they seem fairly feasible in terms of how they would work in the moment.
I totally get your concern about disabilities, and I would absolutely allow any student that have accommodations for technology take advantage of that.
Yeah haha I'm just kind of spitballing ideas as I enjoy doing that. There'd be some pitfalls to it, buuuut then again we've got billionaires selling AI, I think I could sell a college on a digital solution that allows AI and education to somewhat coexist without AI threatening the integrity of the classroom.
I just wish I had the resources to do something with the ideas. haha
I have a couple of anecdotes about this approach.
When I was in college years ago, the new shit was a calculator that could derive and integrate for you. I bought one thinking it was cool, but then I failed calculus not once, but twice. During the second time I was slowly failing it, I ended up leaving it in a classroom on accident and never saw it again. The third time I aced the class because I didn't have anything to bail me out and had to actually learn it; never had a problem with it after that.
When I was working on my pilot's license, my old-timer instructor taught me to do everything by hand. Use a manual flight computer, paper charts, flight plan by hand, that kind of thing. I trained on a plane with steam gauges when the club had plenty of planes with fully digital avionics. The weakest part of my game was flight planning because there were so many digital aids that obfuscated the act of flight planning that I never really learned how to do it. Once I dropped those and went back to pen and paper, I sailed through my exam. My instructor had this motto: learn the hard way so you can graduate to the easy way.
I think it's good advice.
That's great advise, I really like it. Thanks for sharing!
I'm not a teacher or professor but I've been following discussions about the ai problem in the classroom. Some people claim to have substituted oral presentations for research papers.
I think your notebook idea is a fantastic idea. Personally I would have struggled because I have terrible handwriting but for most students I don't see much of a downside.
I actually do have an oral group presentation at the end of this semester students will do that works in tandem with a typed paper. The idea is that I can't really verify whether or not a student has used an LLM to write their paper, and if they do, having to talk about it in front of class will force them to at least read what they wrote.
I imagine that I will struggle a lot too because of terrible handwriting in terms of grading it, but that would be a burden I'd be willing to bear :)
As someone who works writing papers for people, I was scared that the AI revolution and physical paper approaches like the one described here would kill my work. Turns out it is about the same it has always been.
There are AI detection tools, so flat-out copy-pasting does not work, and between fact-checking the AI for hallucinations and rewriting the paper to make it sound human, it ends up being about the same effort as writing the paper yourself, especially for longer 3+ pages papers, so customers continue to hire me. In the case of hand-written papers, the customer simply "hand-prints" my paper with a good old pen.
Therefore, as a means to prevent cheating, this is a waste of the honest students time, and also a waste of valuable professor's time (reading handwriting is significantly slower than reading on a screen)
However, I will note that AI has elevated the quality of my work. You can do so much with it that is not "write my essay". You can ask it for alternate theses, to find obscure facts and data that are hard to find and easy to check, to provide feedback on your paper, to contrast your paper against the rubric, to find factual errors and omissions, to organize your brainstorm into structured paragraphs, and so much else.
If you are smart about it, you can use AI academically and increase the quality of your work while remaining 100% legit since you are not plagiarizing the LLM's output. It's like having a personal teacher. If you are dumb about it, you will copy-paste or replace words instead of properly paraphrasing AI and get found out.
To the professor's credit, studies do show that handwriting can improve retention by forcing deeper processing and involving the motor cortex and other parts of the brain not stimulated by plain typing, so it might genuinely help some students. But mandating it for everyone, to me, feels like punishing the majority to catch a few, especially when smart AI use is already legit and undetectable if done thoughtfully.
I guess the TL;DR of what I am saying would be:
A) Handwriting is ineffective to stop LLM or traditional cheating.
B) It increases workload for everyone for very minor retention benefits.
C) AI can be used responsibly.
So a better policy would be teaching students how to use AI well, rather than being a Luddite about it.
Given the rate of false positives, I can't ethically use AI-detection tools as a proof of a student cheating.
Ultimately, I'm not concerned with students finding a prompt and copying that by hand. This has always been a problem and there is no clear solution for it. But, as you point out, handwriting seems to be the best way of learning and retaining information. If a student is forced to hand write an output from an LLM or someone they hired, they are still processing that information in some way as opposed to copy and pasting or do quick edits to a file (which many students do).
My interest in this, as was illustrated by Ayşe Baltacıoğlu-Brammer, is that I want to improve student engagement with the course. This isn't, and there is no, silver bullet solution to that. Some students will refuse and will find any way to work around it. But if I can have a net improvement with engagement and students take away new studying skills, spend less time on a computer, and think more deeply about the material then I would view that as a success.
As per your last sentence - I already can't cover all the concepts that I think is relevant to the course material. I'm not too interested in carving out more information to dedicate towards responsible and ethical AI use, and on a personal note, I think spending less time on computer screens is probably better for us so I'm OK being a bit of a luddite about this :)
One of the things I took from the article was that it creates a far higher barrier to entry to student's replacing study/learning with the use of AI or hired paper writers (where the student trying to read and understand the output; removing themself from the loop). It is a LOT harder to procure a paper and then copy it out than it is to print off a procured paper, and I suspect that would change whether it is "worth it" to a lot of could go either way students.
I think this is the bigger discussion that needs to happen. Like it or not AI is here to stay, and we do need to start showing how to use it responsibly. The calls to flat out ban it just remind me of the people from my childhood who wouldn't let their kids use computers or screens at all. Like sure, less screen time is good and I get where you are coming from, but a giant portion of the working / real world is computers and screens. It doesn't actually help to hide those things away. I think AI is in that same boat.
I'm not really opposed to the use of LLMs and other tools like it - I acknowledge that they can be very useful and I talk about some uses I am OK with in my course (e.g. a spring board for identifying drug targets to research for a condition, generating names for a drug). However, I also think it's important to demonstrate that you don't need to use these tools and that proper usage of the tools requires a foundation of certain skills/knowledge. We're seeing negative trade-offs in cognitive skills as people offload their thinking to these models, and so I don't think that forcing students to fall back on and develop those skills is going to harm them.
This kind of approach is likely to work well in the short term, particularly for students who over rely on AI to write basic prose for them. However, AI assistance is already opening a gap between "cheaters" and skilled researchers. Smart people need to learn how to critique and marshal AI outputs. If teachers forsake modern tools, their students will be unprepared for the ways those tools will both enable and betray them. The floor is raising; it will become a mistake to also lower the ceiling.
These models have gotten way better in a short amount of time and there are many use cases for them to automate tasks and provide concise summary of material, there's no doubt about that. This may be a lack of imagination, but beyond something like generating a quiz to test your understanding or using it as a smart research tool, I struggle to see how incorporating it into my lessons is going to facilitate the learning process as it pertains to developing their reasoning, which is what I'm interested in. I already tell them to use it as a starting point for identifying drug targets, but that it should be taken with a dose of skepticism because they can't always evaluate the correctness of outputs. I am looking out for workshops to provide some ideas on how to better incorporate LLMs into my class and assessments - I am not completely unwilling to use the technology, but I don't think I should be teaching students how to use it responsibly if I don't have a clear plan for it.
I lack any applicable knowledge in teaching, so this is likely not a useful direction, but may I ask if debating/correcting an LLM could be useful, if your intent is to develop students' critical thinking skills? Not sure how well this could be pulled off, but for example:
For my part, I'm mostly using them to help with learning concepts from recorded lectures in MOOCs: the communities for those have long since died off, and in university, I really leaned on being able to ask for clarifications at the end of each lecture. LLMs have been very helpful in that regard, especially if they cite their sources in their responses (which, as you note, are sometimes a wee bit ... creative).
I do like the idea of turning the student into a tutor, that's a creative idea for a study tool. I hadn't thought about setting a parameter of an LLM of being intentionally wrong - further proof there is a lot I have to learn about them. I was viewing it from the lens of, "They don't yet know the material, so they can't evaluate if the output is wrong". I'll try playing around with ChatGPT and see how I feel about it, thanks for sharing your ideas!
I hope it works out! If it helps, I've found that Claude and Gemini (to a lesser extent) do a better job of responding factually if you're able to upload content for them to pull references out of (and then explicitly tell them to cite from it). I've still had it hallucinate a lot, though, and you really need to urge them to "do research" in order to avoid frustratingly incorrect conclusions ...
(Gemini's current model is extremely bad about this, and gets mouthy if it's corrected ...)