The Effect of Computer-Assisted Learning on Students’ Long-Term Development | Hacker News Reader

The Effect of Computer-Assisted Learning on Students’ Long-Term Development

papers.ssrn.com

79 pointsgandalfgeek45 comments

What I have grown to appreciate recently is the social aspect of human learning. For a lot of us, perhaps most, being in a social environment greatly assists with learning.

Hence having small classrooms where students and teachers can interact with each other.

I found the importance of the social aspect most vividly in webinars:

- A live webinar and a pre recorded webinar where attendees only passively watch and do not participate are exactly the same except for the feeling of being “live”

- That said, individuals are many more times likely to dial into a live webinar than a pre recorded one —- it’s because of the social aspect

I ran a series of webinars at work and saw these principles play out first hand (I did a write up on how to organize effective webinars for anyone interested: https://link.medium.com/Bpq4EH8Ymcb)

I feel we will learn more from an average teacher who we socially interact with than the best teacher from who we are not socially connected. There’s something about humans being social creatures that we best accept knowledge in a social sphere.

I say this as an anti social person who prefers to read textbooks than watch videos or talk to people :D But through the years I realize I am more in the minority.

This is why meetings in companies are very important - for us hackers it sounds like a waste of time - but for a lot of the genera populace, a meeting is an easier way to absorb information because of the social aspects to it. Many people I think struggle to learn anything without an interactive session

A good teacher do so much more than an average teacher. I remember I was in school, and had a few good teachers. They are brilliant. They teach well, you learn faster, and you get hooked with the subject. My chemistry teacher on the other hand was mediocre, I had no interests in chemistry at all, and I did really bad. There were many students in my class did well in Chemistry, and many of us did well in Physics. The difference is teaching. Of course, social aspects is important, when you can have everything, having everything is the best you should aim for.

The study of this shows a solution to schools which couldn't attract good teachers due to location, fund or something else, it is not applicable to general schools and general teaching.

You raise valid points that I agree with. Perhaps to better rephrase, assuming an adequate level of teaching ability, in person teaching or well constructed digital sessions where the social aspect is not removed, is likely more effective than simply listening to the best teachers but in a less social manner.

Our goal should be to increase the baseline level of quality for teachers and not just concentrate teaching to a select few orators who broadcast to large groups.

My original comment was partly in response to another comment on this page, so apologize if it is a bit tangent to the study at hand

> where students and teachers can interact with each other.

There's another dimension to interactivity that is not merely social.

Really excellent teachers in a "small enough" class-size are able to assess individual students IN THE CLASSROOM. Interactivity in this context means the teachers pose questions to students individually or as a group and orchestrate coherent learning-focused discussions. This is the basis of the Harkness Method which is used in the most elite private schools where class-size is in the singe digits.

Good point (and also hard to achieve - I tip my hat off to all the excellent teachers who do this everyday, not an easy task to achieve).

Interestingly I've made a successful career in this field in the opposite way it sounds like you did. I dropped out of college for unforseen circumstances and had to self-learn things like maths, data structures, and algorithms as academics routinely use these as litmus tests in interviews.

My learning wasn't social. It was spent during my breaks at my dead end jobs, after work, etc. If I found a topic I didn't understand there was no tutoring, professor to reexplain, or even friends to ask (for a long time I didn't know many people in this industry)

The result is that I learn new things quite easily, I can usually distill topics down easily for folks or represent them in a multitude of ways, I'm unafraid of failure, and I don't spend my workdays socializing. I still enjoy socializing, just not at work.

This is why meetings at companies seem so frivolous. What could be communicated in an email, slack message, or hallway conversation is left to a meeting and I often wondered if it's just because these folks want free lunch and some socialization.

I had the same experience. Exactly the same thoughts about endless meetings about stuff that can be easily sent in a message...

Just playing the devil’s advocate here based on the top comment but what do we really mean when we say this meeting could be an email?

Perhaps, the meeting isn’t interactive enough? Why have a meeting at all if not everyone in the meeting is providing input/asking questions? If the topic is already done and decided and the chapter is closed, I agree it could be an email to a list and you can reach out to the presenter with questions.

Things like when a boss calls a meeting to say we are working from home until further notice and nobody is allowed to talk. But if we allow people to raise their hand and ask questions and things can change based on participant input then maybe a meeting is warranted? Just thinking out loud...

It generally means that all of the signal amidst the noise and confusion that took up 30 minutes to an hour, multiplied across all the participants, could have been distilled down into 3-5 bullet points.

Most meetings are held to transfer information, in the most lossy format possible.

Is it though? It may be lossy to you because you are better and absorbing information through other means, however my point is that the social aspect is important for many to absorb information, even though it looks inefficient.

For example, many people are sheep who follow consensus —- meetings give an opportunity for them to establish consensus and hence accept and absorb it.

Note I am playing devils advocate here, I hate meetings but I think I get their role now

I think bosses like meetings because they're not as interested, sometimes, in people following the edicts as they are in people doing as _they_ say. It's important for them to be the one telling you and you having to obey. If you just did what an email says then that's just working for the business; the machine is working. If you attend a meeting then, it appears at least, the boss is the one who controls you rather than your actions being directed to the business needs you're [apparently] working for the boss. They say jump, you say how high.

Maybe it's minor megalomania. IME that's another reason for meetings.

Curious, are you a boss, or have you ever been?

Have you ever worked with humans? Cause I guarantee you that if you attempt to communicate via one way email only, your chance for failure will go super high. You will be unlikely to succeed.

You know, our team at multiple points eliminated meetings almost entirely. And in another team, young boss was so afraid to not waste time, that only one person (manager/analyst) was at requirements gathering meeting.

In the first case, the team became dysfunctional each time it happened after like two months or so. In second case we were unable to make customer happy, because it turned out that a lot of information got lost. And programmers were unable to make good decisions.

Maybe the person you are responding to is not better then everyone. Because meetings are rarely about one way communication. Out is two way and lack of those back and forth and fights costs. And where meetings are about one way information, it a.) puts priority on it like this is more important b.) likely puts away people time to absorbed information. Because prior don't read long emails not just because they are stupid lazy, but because they have other priorities.

There are useless meetings too, but once you kill them all you will find out some of them were necessary.

Agreed with all your points

But I think you are of well above average intelligence and also more suited to the study methods you described.

Most people I think want to learn via social environments, if you try and see from their perspective it will make your meetings go smoother IMO - think that they struggle to absorb information without the socialization aspect, and then you will start to realize the importance of some of these social constructs like meetings and their place in learning.

> I often wondered if it's just because these folks want free lunch and some socialization.

In over 50% of the careers I have had in the US I can honestly say that it was more about the ceremony of "work" that it was ever about the work. So much of this is the middle-management dilemma where as a producer/IC I find myself more worried about the outward appearance of my work vs. the actual substance of it.

> I don't spend my workdays socializing

I'd advise against this in most places. I've done the whole "work not socialize" thing and it comes up in 1:1's as anti-social behavior. I'm not saying I disagree with you _at all_ but I've learned to go through the motions and begrudgingly participate in "dev beers" at 2:30pm on a Friday, or Go-karting, or the lavish meals... I've had to go to dance clubs for my job, I've been pressured to go to strip clubs (never went), I've been pushed to perform for talent shows... It infuriates me. It makes me mad enough to punch a wall. I am here to work. My culture is work. Let me work. I am a god-damned engineer - let me perform the work I am paid to do, then let me go home afterward...

Socializing is largely the new American work culture. It's pathetic, and it's a bubble that will pop. We didn't get to the god-damned moon knocking off at 2:30pm on a Friday for "dev beers"... =|

---

Also, as a drop-out I totally feel you. Self-learning was the only way to getting a successful career after a very toxic "education" experience I went through in the middle-of-nowhere Iowa. I legitimately do not know if I would be OK if it wasn't for the internet. It allowed me to self-learn whenever I could, and allowed me a platform/audience to hone skills.

> I'd advise against this in most places. I've done the whole "work not socialize" thing and it comes up in 1:1's as anti-social behavior.

Same. I've had managers in the past express this in different ways. Eventually I learned to just be present, smile, and crack some jokes. Usually that affords me the ability to say my dog needs me, or something.

> Self-learning was the only way to getting a successful career after a very toxic "education" experience I went through in the middle-of-nowhere Iowa.

Mine was a private school. Interestingly the content was great. The methodology for teaching was not so much.

> I legitimately do not know if I would be OK if it wasn't for the internet. It allowed me to self-learn whenever I could, and allowed me a platform/audience to hone skills.

I thank my lucky stars the internet and IRC knew what to do with me.

Thanks for sharing, we should definitely get weekend-citizen-beers sometime.

> This is why meetings at companies seem so frivolous. What could be communicated in an email, slack message, or hallway conversation is left to a meeting and I often wondered if it's just because these folks want free lunch and some socialization.

One thing that remote work during the pandemic has demonstrated to me is the line at which chat rooms lose their efficacy as a tool for collaborative problem solving.

Many times a week I need to upgrade a conversation happening over chat to a video conversation, simply because The linearity of chat sessions makes it quite difficult to communicate more complex concepts, especially causality. Something about spoken language allows us to communicate more dimension.

Furthermore, regularly scheduled face-to-face meetings over video are the only times I get to hear about work by members of my team whose day-to-day tasks do not overlap with mine.

It's not just for the purposes of socialization - though socialization is very important in any team - but rather very frequently in meetings I learn about something that someone else is working on that is consequential to my own work.

> I feel we will learn more from an average teacher who we socially interact with than the best teacher from who we are not socially connected.

Further - I wonder if students would do much better with 1/3 the amount of school time and 1/3 the class size. Instead of 6 hours a day in a class of 21 kids, 2 hours a day in a class of 7.

It’s a good point - there may be a point of optimization, however my gut feel is 2 hours is too short and 6 hours too long. But that’s pure speculation as it’s not my field

As a parent of two young children, 6 hours at school seems short.

If both parents are away for 9 hours -- 8:30 AM - 5:30 PM, work 9-5 and half hour commute each way -- then who watches children for the 3 hours not at school?

China's school hours are 7:30 AM to 5 PM, and include a 2-hour lunch [1]. That seems to yield academic excellence and a strong economy.

[1] https://www.infoplease.com/world/social-statistics/school-ye...

So we're optimizing for parents and their childcare needs. It always seemed to me that childcare was a separate but intersectional problem and we do a disservice to our students by confusing it as the same problem. Also China feels like a very poor comparison target as a westerner

Childcare and education are not the same, but for working parents the intersection is massive and unavoidable.

My question, phrased differently: given a scenario where both parents are away at work for 9 hours per day, and school is 6 hours per day, are the remaining 3 hours not in school more enriching for children than time in school?

Maybe it's better for students to have 2 hours of instruction and then 6 hours of studying with peers/clubs/etc than 8 hours of traditional schooling.

School where my kids are is doing lessons before lunch and then kids play for hours till parents come.

It is absolutely awesome that mothers are able to work. Not just on terms of having money and independence, but also in terms of not having own ego completely dependent on kids. And especially when dad is death or absent, it makes super massive difference.

> we do a disservice to our students by confusing it as the same problem

Absolutely.

The simplest approach is to specify some hours for learning, starting at 9 AM, ending somewhere between 2PM and 5 PM depending on age. However, the school provides optional childcare starting at 6 AM and ending at 6 PM, where the parents can bring/take their child at any moment. So parents bring their kids to the playground whenever is most convenient for them; at 9 AM all kids are sent to classes; after classes kids are sent to playground again; and then parents pick them up whenever is most convenient for them.

When I see schools that try to implement this, the greatest mistake is when they insist that the morning and evening childcare needs to be provided by the same teachers who provide the education. And the teachers of course object, not only because this reduces their status from providers of knowledge to mere babysitters, but also because they want to be with their families. (Then the school says "dear parents, we honestly tried, but the teachers are uncooperative".) Instead, the school should hire extra babysitters for those morning and evening hours. For example, retired teachers would probably welcome the possibility to make little extra money in the mornings or evenings, in a familiar environment.

The babysitters can be fewer than teachers, because there is no need to separate kids by grades. For example, you could start with one babysitter at 6 AM, joined by another one or two at 7 AM, perhaps joined by a volunteer teacher at 8 AM; and then at 9 AM the babysitting is over. Similarly in the afternoon.

I think 2 hours would probably be plenty. I've seen schools during the pandemic have about 1 hour of teacher/student of subject instruction, and that's with larger cohort sizes (9-10 students) and doing everything virtual. You can also look at homeschool examples, here's one I just randomly found[1]. She spends about 1 hour and 15 minutes actively teaching her 5th grader and 30 minutes with her 4 year old (they also study on their own or sit in with each others classes).

As others mentioned, a big part of the issue is that school ends up serving as a daycare for many parents. It's likely that school isn't optimized for education, and there is a good chance that much of school is just for occupying kids that we as a society aren't willing to accommodate in any other way.

[1] https://tablelifeblog.com/typical-homeschool-day/

Afaik, no. There are diminishing returns in making class sizes super small.

There are lots of factors at play here, but it's interesting to note the student-teacher ratios the paper describes. The under-performing Chinese rural schools are at ~17, and they say that urban Chinese schools are at ~12. 17 is roughly typical student teacher ratio in US public schools, regardless of status. In the US a ratio of 12 or less is common in elite private schools.

I was expecting that the CAL program they implemented was what is known as an "Inverted Classroom". It's a learning system where students view recorded (or live) lectures by master teachers as "homework", and then use classroom time with a regular teacher to practice/drill/explore what they've learned-- not unlike a college recitation session with a TA.

But that was not the case in the Chinese CAL study. Instead...

> In our context, however, remote learning happened in the classroom under the direct supervision of local teachers, therefore limiting distractions and procrastination.

Yep - from this I take it to be a more unique model that doesn’t sacrifice as much of the social aspects of learning, as the inverted classroom would where there is a feeling of detachment during the viewing of the recordings.

It seems silly to have 1,000 teachers who range from first best to 1,000th best all teaching roughly the same number of students a year. Why not give the top 10 best teachers all the resources they need then let them teach everyone?

Monoculture? I don’t think teaching is binary. There are of course universally really awesome teachers, as well as terrible ones. But between them, there are all kinds of teachers with all kinds of teaching styles and crazy experiments. It may not be equal for everyone, but I think it’s much better than monoculture.

Then include heterogeneity of approach in your ranking system. Even if we do this we still have a rank order from which we can select the N best and give them a scalable platform through which they can reach large numbers, and be better off than under the archaic approach of a bijective mapping from random teacher X to Y students (Y less than 200), where the quality of X is probably dominated in a Pareto sense by any single teacher in the top N.

Caveat here that I'm no expert in pedagogy, but I would also say that this fits well with the part of the teaching process that can be scaled with no downsides, eg lectures. But TA sessions etc can't be scaled as well and such an approach wouldn't fit.

Just a little empiricism: During my studies, 95 % of the teaching was somewhere between "not good" and "horrible / comically bad".

There were just one or two persons in the whole faculty known for good teaching. This was acknowledged by all students I came in touch with.

--> There is a kind of teaching that is uniquely perceived as "good" among the people I came in touch with.

The same holds for teachers known for really bad teaching.

That written, teaching _was_ mostly perceived binary.

There was fairly little overlap where teaching styles were perceived "good" from some and "bad" from others.

I think that's like asking why nine women can't make a baby in a month.

Who defines best? Why not give all the resources to one teacher?

Students (or their families) could choose the definition of best they wanted. Some teachers are better at adding value for less intelligent pupils, some better at pushing the more gifted, some best for the hardworking, some best for the ill-motivated ... take your pick. It can even come down to relatively inane factors like one teacher's diction being too difficult for some to follow but fine for others.

Re your 2nd question: bus factor. And the same answer as the first - that you can have different definitions of best depending on the pupil.

Why not the top 1?

I don’t know tbh. Feels wrong though.

Firstly, bus factor. Then robustness. How do you train new teachers? How do they get experience? Then theres diversity of thought, where do you get that from if everyone has the same progenitor of thoughts?

Then what happens politically when you give such a few number of people such power. Would fundamentally change politics.

What abuses of power? Let the wrong person into those few teachers and give them lots of control over millions of children's day to day lives? Thats a lot of children to screw up if you get it wrong.

That aside, it goes against my personal belief about how the world and humans work so I can’t get down with it.

Not at all. One of the major concerns in schools is student to teacher ratio and small classroom sizes are considered a plus.

All the well-meaning but naive plans to turn education into mass media would with up with all the wealthy zip codes having teachers and everyone else relying on glorified youtube.

At this point, if there’s an education problem where technology would be useful, it’s probably something subtle like collaborating to publish an affordable textbook as a way to break monopoly pricing.

Good teaching of kids is not lecture - relationship and reacting to kids as individuals matter and does not scale.

"which connected China’s best teachers "... I wonder why they choose the best teachers for this type study? Wouldn't a randomized, average sample produce more objective results?

I am extremely dubious that they can properly control for other variables to so blithely offer up "Computer-Assisted Learning" as the cause of any changes, even for values of "Computer-Assisted Learning" limited to "State-sanctioned and enforced computer-assisted learning with novel infrastructure and tools from a world-class expert as experienced by a culturally and socially uniform population of rural Han Chinese students studied for only four years."

And, of course, I wonder if recent behaviour by the PRC should encourage us to be confident that reports about the success of multi-billion dollar government initiatives should be considered entirely trustworthy.

> Second, the program installed 264,905 satellite receiving sets, comprising satellite antennas, satellite TV equipment, computers, and other related devices.

No internet in a region -> internet is a huge important jump.

I would expect of A) DVD's B) DVD's and Internet C) DVD's and Internet and Computer Rooms

B) Would win.

C) Would get into the Wests issue of installing useless tech that just distracts students, even in rural China in 2004.

I can't find if they compare these installed models?

to be clear, the "assistance" the computer is providing is to connect rural learners to teachers. the title is a bit ambiguous

Reply on news.ycombinator.com