Nobody Knows How Well Homework Works | Hacker News Reader

Nobody Knows How Well Homework Works

astralcodexten.substack.com

39 pointsfeross38 comments

In my early teens I was really frustrated by the amount of homework I had to do and how much time I felt it stole away from my outside interests and hobbies.

Looking back, I was right to think that way because my entire career has been built on things I taught myself via those hobbies and interests. Not even indirectly. Completely directly linked.

Even thinking objectively about the meta skills that “homework” may teach you, I still think it has accounted for 5-10% of my useful professional/life skills vs my own adolescent interests (programming, electronics, music production and entrepreneurial stuff) which I peg at 90-95%.

If anything, the only thing homework achieves is to make children used to being forced to do something for the sake of doing it.

People learn what they want to learn, children that aren't interested in the material can finish homework assignments but not retain anything.

Teaches children to triage what they're required to do, to get the bare minimum done without hurting themselves too badly.

I had the same experience in university. It felt like the workload got in the way of learning the stuff I was here to learn. I had no time to be curious.

100%, and I'd go even further.

The time I spent playing video games was better for my development all round than my time spent doing homework. The time I spent reading was far better. The time I spent outside playing was likely even better again, in my particular case.

I question the legitimacy of a system which teaches the scientific method yet kills millions of years of child development with utterly unproven busy work.

> I question the legitimacy of a system which teaches the scientific method yet kills millions of years of child development with utterly unproven busy work.

Yeah this is a very good way to put it

There is even an xkcd for this https://xkcd.com/519/

On the other hand, the homework i hated the most was writing essays for english class. To this day i still kind of hate that sort of thing. On the other hand, that type of writing is very similar to the type of writing they ask you to do in "peer reviews" during performance evaluations. Who knows, maybe if i was better at it, it would help my career.

That XKCD is surprisingly spot on - in my case it was quite literally a weekend spent learning Perl that kicked off my early web development career!

homework teaches obedience, submit or get punished

I rarely had to do homework outside of school. Typically I’d just do it during the class.

This is a fairly complex topic. In my youth, I certainly didn't enjoy homework. Even in subjects I was passionate about. I much rather study the topics that interested me in my own way and pace than doing assigned problems from a textbook. Even if that meant doing all the problems from the textbook lol

More recently I discovered that the school I went to abolished homework. Kids there are supposed to complete all school activities within the allotted school time only.

However, homework may be valuable in the case where the student would otherwise not study on their own...

It's complicated further now by gamification. My daughter's school doesn't assign homework until a certain grade, and minimally at all. But they're encouraged to use many apps on different subjects that are structured as games, which she really likes.

Reading this post, I realized it's very fuzzy whether she's doing homework: she's not doing it in the traditional sense, but if you include these apps, which cover all the same content, she's doing them all the time.

One topic not discussed here, is that the function of homework is not only to increase attainment in the topic. I work hard. I strongly believe my hard work is a significant factor in my success. I didn't always work hard. In my GCSE exams (16 year olds UK Exams) I got a mixed bag of results, I didn't particularly work very hard. At AS level (17 year olds) I kind of realised that if I didn't start working hard I was not going to get good results. I started working seriously hard, particularly self-teaching the parts of my subjects that were particularly poorly taught by my school. I went on to get straight A grades (highest grade available at the time) in both my AS levels, A levels, and a first class masters degree from a top university. The role of homework can be to teach children the value of hard work and to teach them the skills and habits that are needed to be successful adults, that is more about culture than what grade you actually got in your middle-school algebra class. These skills and attitudes are also very difficult to pick up later in life.

Homework taught me to physically suffer and degrade myself to the extent I couldn't avoid it. Can't imagine how it encourages positive work ethic.

Finland assigns the least homework, and gets the best educational outcomes.

https://in-finland.education/homework-in-finland-school/

While "the success of education system in Finland is provided by many factors, starting from poverty rates in the country to parental leave policies to the availability of preschools", the above still seems pretty cut and dry.

I think the problem highlighted by OP is that we don't have proper data on homework.

At best, the stats you share show that 'it is possible to deliver the best global educational outcomes without giving children homework'.

We don't have proper data on whether these educational outcomes could be further improved by assigning more homework.

I think Finland is not a particularly helpful example to share here, specifically because they're widely lauded for having an excellent educational system across the board, they could quite easily still be world-leading even if they've got it wrong on homework (although I accept it probably means that the best education researchers have likely considered the homework question in Finland and have come to a position on it).

From a personal perspective, I think there are a lot of factors to whether homework is useful.

For example, in a former life I was a languages teacher, and I saw the children I worked with for one or two hours a week. In order to learn vocabulary, without a doubt the most effective method I've ever used is spaced repetition. My best chance of teaching vocab items is to open the lesson by introducing them, a mini-recap at the end of the lesson (maybe a memory quiz, etc.) and then for the children to review themselves at some point before the next lesson. I would imagine similar techniques would be useful in most subjects.

Anecdotally, I saw faster and more secure progress from those children who had done the homework.

Additionally, those children who had done the homework tended to be better behaved, more responsive to 'course-correcting' when they got distracted, and were able to exhibit better self-control... obviously all of this carries a heavy caveat that this may be WHY they were more compliant with the homework, rather than a consequence of it.

However, with a bit of common sense (and to be clear, I have not searched for evidence for the below, and it may well not exist), it seems to me:

1. There are some things where short stints of 'self-learning' are clearly beneficial, and it is either not possible or not an effective use of time to do this during teacher-led lessons (eg: vocabulary/rote learning, independent reading of literature, etc.).

2. A requirement to do some task which does not lead to immediate gratification builds a habit of doing things that are not necessarily fun just because they need to be done

3. Learning to be able to self-manage and complete tasks by a deadline is a skill to develop on its own

4. Learning how to cope with not having completed a task just before the deadline is also an important skill ('can I get away with very quickly doing this in the break before the lesson? can I copy from someone else? should I admit to the teacher that I haven't done it? should I make up an excuse?' etc.) - as long as it's relatively rare

5. Excessive amounts of homework which interfere with a child's private life/free time/etc. are unlikely to be helpful - between 0-5 hours a week total across all subjects seems like a fair amount.

6. Schools should take care to co-ordinate homework so that it's balanced throughout the week (eg: giving 3 hours of homework to be completed on Monday night and none for the rest of the week is not sensible).

7. Teachers should not use homework as an excuse for not planning time properly in a lesson ('oh we're out of time, please finish this at home')

NB: Edited for formatting.

You make some fair points on homework's fringe benefits, but let's be candid here: None of that is why schools are issuing homework in the vast majority of cases. There are other paths to teaching autonomy and delayed gratification, which don't use the most tedious and ineffective way to do so.

There's nothing sensible about throwing millions of hours into homework without any evidence that it's actually in any way a constructive use of time. There's more evidence for positive benefits of video games.

I don't just doubt that homework is a sub-optimal use of time, I doubt that it's useful at all in at least 90+% of current real life use cases. That ought to horrify teachers, and it doesn't, and that freaks me out a little tbh.

Spaced repetition is great, and I'm glad you mention it. Multi modal learning has great evidence behind it too. There are many, many more activities, strategies and methods that produce strong and consistent results - yet are only used by rare individual teachers (the ones who "do their homework"). Conversely, homework [by which I mean the pointless busy work we all know and detest], has virtually no evidence of benefit; practically none.

... So why would anyone put *any* stock in a system that ignores SO MANY proven learning strategies, in favor of something which most children find utterly detestable, and doesn't even work that well?

Like, on balance, I would bet any money that homework is net negative. It kills childhood, it kills love of learning, and for no provable tangible benefit other than molding adults who accept what's becoming known as "bullshit jobs".

I don't think they're 'fringe benefits', but I do agree that for the most part, schools are not assigning homework for these reasons. I also absolutely agree that most homework is assigned primarily as 'busy-work', because the teacher feels that 'children must do homework', rather than because they're trying to achieve a specific educational outcome, and this is the problem.

Fundamentally, I think it's better to give no homework than bad homework... but better to give good homework than no homework, and setting good homework is much more difficult for the teacher than setting bad homework - so your conclusion that 'homework is net negative' is probably true on average, but is not necessarily true for any individual child, class, or school - especially if they attend an excellent school with well-informed and caring practitioners who use homework thoughtfully as part of their pedagogical methodology.

I think we're both arguing two sides of the same coin though, and we're in agreement that the majority of homework as it is set today is not worth the paper it's written on, and although I perhaps wouldn't be quite as emphatic as saying it 'kills childhood', I agree with the overall point that most people's lives would be better if they a) didn't have to do homework if they're still school-age or b) hadn't had to do homework if they are no longer at school.

AFAIK New Zealand doesn't do homework either, only reading.

Homework, as imperfect as it can be, is still practice. There's always value in practicing something you learned, preferably by being challenged to answer questions you're not used to.

Opportunity costs are a thing.

Making children hate school is also a real risk.

Homework is what someone overworked would come up with to force everyone, wether they need it or not, to put X amount of time repeating and practicing Y.

I think it's a frustrating experience overall and we should be looking at giving students the entire curriculum from elementary to masters degree upfront and letting them work up from it at their rythm.

As incentive students should be provided X years of access to teachers free of charge, and if you finish early pocket the difference. If you finish late you have to pay the difference.

But that would mainly benefit kids from educated backgrounds, while kids who graduate late because of their life circumstances, might end up simply not finishing school or start their professional life in debt.

After reading some of the school- and homework-critic posts here and thinking I‘d agree with them I realize that school helps the overall population, especially less educated people who simply wouldn‘t invest more their time and energy in education otherwise while the highly curious and intelligent people are rather limited by it.

This is one of many many things about the American education system that is completely done without any scientific proof of efficacy. Often when there is data, the data proves it doesn't work, but we "have always done it this way". Our education system is shockingly out of step with modern life, yet we do it this way simply through inertia. Consider our system pretty much hasn't changed since the industrial revolution, and was designed to produce factory drones.

See New York state teacher of the year John Gatto for more: https://www.cantrip.org/gatto.html

"Look again at the six lessons of school. This is training for permanent underclasses, people who are to be deprived forever of finding the center of their own special genius. And it is training shaken loose from its original logic: to regulate the poor. Since the 1920s the growth of the well-articulated school bureaucracy, and the less visible growth of a horde of industries that profit from schooling exactly as it is, have enlarged schooling's original grasp to seize the sons and daughters of the middle class. "

In college I had a class where 25% of the final grade was Exam One, 25% Exam Two, 25% Exam Three, and 25% Homework. I got 100s on all of the exams, but never did any homework, so I got a C for the class.

In later semesters, I realized I could convince professors to give me a higher letter grade than what the sylabus prescribed, but for that class, oh well.

For doing well on exams but not the homework? If so that's very lenient of them, but understandable. I prefer professors just make homework extra credit and make exams a large part of the grade.

But you prepped for the exams, right? That's the equivalent of homework.

If not, then your situation is an extreme outlier.

There are many types of homework and undoubtedly some are not useful. These fall into the category which we often refer to as "busywork" - they are assigned for the sake of assigning homework because that's what instructors feel they should do.

The useful kind of homework we can categorize as being helpful for the exam, and those which are useful even beyond the exam. In principle, they should be correlated, but there are also many ways to create useless exams too.

Be honest. How often would you expect to do zero homework assignments throughout a course and get straight 100s on all of the exams?

That's the parent's claim that they did no HW and got 100s on the exams.

Debating whether different HW assignments have varying ROI is a different issue.

No. I didn't study at all and graduated from a major state university with an engineering degree. If I had to do it all over again, I would take it more seriously.

(I am not an engineer. I just thought an engineering degree would be the most practical to take/learn. I could have learned more, but at the time, I didn't have the desire to.)

How far back is this??

College in the 90s, I had very little homework outside of papers and programming assignments....

Around 2010

In my opinion, homework is first and foremost should be about two things: the first one is being able to problem solve, research and study on your own (developing your own "style" of studying so to speak) and being able to apply the studied knowledge outside of spaces that you're used to apply the knowledge at (classrooms). Given that remote learning and homeschooling are going to only increase in popularity, i wonder if blurring the line between the actual "classwork" and "homework" would have any sort of a positive effect on applying and retaining the knowledge.

That's one benefit of homework, but the thing is I've never been really taught skills for problem solving or how to study. For example, most students just reread in order to study for exams but that's been found to be the least helpful method.

Shouldn't school teach study methods and problem solving strategies first, before assigning homework?

That's a good question. I think that you're supposed to "develop" those strategies on your own as you attempt to resolve a problem, but i don't think that it works as well as everyone expects it to because when you're young you're always going to have something more interesting to do than some task that is associated with school (skewl sux1!!), meaning that instead of getting any meaningful result, you are going to attempt to accomplish the task as quickly as possible, so you can do what you want.

The dilemma of the school system is: if you try to go by the book, provide as much knowledge as possible in the shortest amounts of time – you're going to have to resort to lots and lots of homework. That isn't going to be welcomed by many (because Billy read on the Internet that skewl is ackshually made so ur gonna work at like...uhhh factory and...and...uhhh...homework SUCKS). You could also abolish all the homework, resort to gamification, teach solely the basics (and don't forget da taxes (and who cares if taxation differs by the state/industry)) but at that point what would the point of school be then?

I don't think that describes any of the homework i got at the high school or middle school level.

Homework has always been extremely helpful for me. I learned a lot of the harder concepts that require more time and thought via homework. But this depends on good teachers curating the right problems/questions. To each their own I guess

I go from the article back to the HN comments, and discover that in fact Everybody knows how well homework works. Scott Alexander just asked the wrong people.

My impression is that it depends. I have got myself unstuck by doing every damn exercise in the book, checking the ones with solutions in the back, and seeing where I went wrong if I did. I have also seen terrible homework that was simply a time sink.

Reply on news.ycombinator.com