Step 1: give every kid a laptop. Step 2: learning begins? | Hacker News Reader

Step 1: give every kid a laptop. Step 2: learning begins?

arstechnica.com

40 pointsllambda37 comments

Teaching is a bizarrely evidence-free field. It's as if there's an active aversion to finding out whether a method or tool works or not. I can't think of any field where so little is known, where there is such a paucity of good data.

We've been spending hundreds of millions on putting technology in classrooms, without a shred of evidence that it improves learning outcomes. As a society we do plenty of things that we know not to work, like imprisoning people or criminalising drug users, but at least we've got enough data to say that they don't work.

That's because the goals of teaching are bizarrely ill-defined. With drug laws, we have a concrete, measurable outcome we're looking at. Most attempts to actually define a goal for teaching (e.g. with standardized tests) have been derided as measuring the wrong things and harming education in the process by forcing teachers to "teach to the test".

Spot-on. It's incredibly difficult to measure intelligence, and learning could be defined as the delta of intelligence on a particular subject. The best way I can think of would be real-world application of the knowledge (in the form of applied projects), but that gets to be too time consuming and expensive in a world where "education" is mandatory to the point of being a commodity.

Exactly.

Current efforts at educator performance are attempts to apply basic economics to a real-world complex multivariate issue - how does a given teacher affect the arc of a child's overall education, when decoupled from the child's home environment and overall societal trends?

Accurately measuring this kind of impact would be more similar to modeling weather than say, worker performance. A more accurate model would be like lifetime patient medical outcomes (which we also are poor at measuring in detail)... if we could measure how good a doctor is by the healthfulness of their patients, then we have the basic model by which we can measure educators.

I think part of the problem is that it's a very difficult question to quantify - we can't control the inputs, we're not totally sure what we want from the outputs, it's unlikely that a uniform process will work to get the outputs we want, and we don't have the materials to guarantee a uniform process from start to finish.

On the input side, each incoming class is different: they vary across geographic and socioeconomic lines, they're all affected by factors which occurred years before we ever got them in a classroom to begin with.

On the output side, we really _Don't_ know what we want. It's easy to say, "students who can read, write, and do math," but the devil, as always, is in the details - how do we measure that in a way that doesn't immediately become a metrics problem á la No Child Left Behind?. More than anything, we need to foster successful individuals - but we're not sure how to do that on a large scale and the education system is but one part of a child's life.

The answer in a lot of ways seems to be "Very Good Teachers" - but what makes a good teacher? What sort of support structure is required? Are we willing to pay for enough Very Good Teachers? Do we have enough? Can we train more?

Further, to what degree does the attempt to constrain the education system so we can quantify the results inhibit individuality? In assuring a child knows his multiplication tables, are we imposing a system that will prevent that child from the free expression that would eventually lead to a great artist? Many of our finest works - the pieces of art by which we define our culture, spring from people who would, by any other metric, be considered failures by a system intended to raise successful individuals.

All this is complicated by the fact that the end result of a good education system ultimately is a good society - something we won't know until 20, 30, 40 years later. Everything else is a proxy: We think a high college attendance or graduation rate will get us there, we think more scientists and engineers will get us there, we think more well-rounded individuals will get us there, but ultimately, we don't know what makes a good society, nor what works to get us there.

I think there are improvements we can make, teaching methods that work better than others, and probably ways to quantify some of those differences, but ultimately, we don't even know the bounds of the problem we're trying to solve in education.

Great comment. I think the difficulty in defining what makes a good teacher is as difficult as what defines a good parent. In rough terms, I think a good parent means someone who raises a compassionate, honest child.

Similarly, a good teacher / education system is one that raises inquisitive students. I've long believed it's better to graduate kids who love algebra vs. those that know and hate calculus. Because in 10 years, the ones who loved algebra will probably end up loving calculus too (instead of developing a lifelong aversion to math, and learning in general).

So, my meta-reply would be to search for a system that tries to generate genuine interest in learning. (Plug: I try to blog in this manner at http://betterexplained.com).

In my high school we were all given laptops and all the teachers had SmartBoards. Here's how the technology was used:

1. We all gamed in class and got a shred of work done.

2. A fraction of the teachers used the boards, even though they were all trained how to.

3. They were crappy Dells that spent a good deal of time being repaired so many students were without a laptop on any given day.

4. I was personally taken home in a police car by the school officer and had my laptop confiscated for using VNC on a teacher's unsecured Smartboard. Those things caused way more trouble than good.

Oh...and our rival school all bombarded the local Blackboard servers on Exam day and DDoS'ed their way out of exams.

Of course, quite a lot is known about what makes a good teacher. It just isn't widely known, let alone taught to new teachers, let alone to current teachers.

See http://www.nytimes.com/2010/03/07/magazine/07Teachers-t.html... and the book that came out of the research featured in the article: "Teach Like a Champion".

Laptops and tablets are the most distraction-prone pieces of technology yet invented. That anyone would expect dropping them into the hands of children to increase their consumption of traditional rote-learning curricula is just bizarre.

The article's point about basic infrastructure is right on. What allows me to really concentrate and be productive with my $n-thousands of dollars in computing hw/sw is that I'm in a relatively comfortable, relatively calm environment with easy access to a bathroom and snacks.

[That, and I've hacked my way into responsibilities that are adventure-like. Totally agree with fellow-commenter's reference to The Diamond Age.]

Take those things away, and the primary use for my laptop becomes either a) distracting myself from the fact that I'm uncomfortable, or b) using the laptop to procure some combination of peace, bathrooms, and snacks.

Startup idea: "like Foodler, only for toilets and earplugs!"

I could understand a project to give every child a (good quality) e-ink reader. These devices could be semi-locked down, and have minimal specs. That'd reduce the opportunities for use other than reading text books.

People can highlight passages they're finding difficult. There could be basic exercise questions (and these could even be web enabled).

It'd fit in well with open source text books; allowing updating of content when needed, with corrections and improvements over time.

There are a few drawbacks - there's very little interactivity and no chance for video. People with dyslexia aren't going to get best use of these devices, nor are people with learning disabilities or visual impairments etc.

But the cost would be very much cheaper than the current dead-tree text book system.

Are you referring to the in the US? If that's the case I think Universities are starting to do that type of thing and it may filter into high schools in the next couple of years.

Forget the hardware.

What we really need is a way to get Anki (spaced repetition learning software) hooked up with enough logic (AI like?) to allow individualized learning with different instruction methods that tailor themselves to each student based on subject.

Of course, what I am really saying is that Neal Stephenson showed us the way with "The Young Lady's Illustrated Primer" in The Diamond Age.

Now all we have to do is make it.

Isn't that basically what Khan academy is trying to do?

Maybe. They are certainly aware of spaced repetition.

I am not sure what form any adaptive response the exercises take. Any fellow HN'ers with insight?

This is precisely what Kumon learning centers already do.

I worked at one for a few years as a tutor. Annoyingly, the biggest complaint from new parents was "My kid already did this worksheet! Why are you giving it to him again?"

I had to explain that repetition was the most important part. You could explain the science but the parents don't seem to care about that part. The parents often thought that it was de-motivational and belittling to have to do the same work over again, so we often mentioned that revisiting something you've done before and acing it the second time around can actually boost confidence quite a bit (we would sometimes make a point of showing a kid that he might not have done so well with something a few months ago, and now he did it perfectly or nearly perfectly).

PS: 99.9999% of tutoring success depended on motivation. I could teach any kid in the world anything if he was motivated, regardless of his or her knowledge when they came to us. There were only two things I figured out about how to get kids motivated that weren't:

1. Enthusiasm is contagious. The tutor damn well better be excited about the subject. The environment better be positive, and sometimes that meant telling the parents to back off. Most parents I worked with were Indian/Chinese/strict. Most were extremely good and supportive, but some were really harsh.

2. Provable, tangible progress that you can show the kid. This is another reason repetition is so important, it's half of the motivation, even if the kid doesn't outwardly admit it.

I'm taking a class through my university right now that uses LearnSmart from McGraw-Hill. I was pretty irritated having to buy a subscription to a website in order to do my homework ($50 plus the $230 book, seriously?), but they have a study guide that you have to go through to get to the homework that is actually quite helpful. Right up front, it presents you with the question then asks how sure you are that you know it. If you don't know it very well (even if you get it right) it keeps asking the question until its sure you know it. You're supposed to go through this every day (or so), and it will ask different questions mixed in with the older questions.

Worth the $50? That's debatable. Helpful? You bet your sweet ass. When I miss a question then get it right the next time it comes around, it's difficult to describe how I feel.

I think it's pretty obvious the Maine paradigm is the way to go. I think peter_l_downs (1) and GigabyteCoin (2) and are on to something: games get boring. Video is the biggest problem.

I have two kids, 10 and 7, and a slightly obscene number of computers.

* They each have OLPC laptops, which they have no interest in. Occasionally, I'll dig one out to show them how they can manipulate the programs, but they quickly tire. No news there.

* We have a Touchpad and an iPad.

The TouchPad charges in the master bathroom, and is basically an extremely fancy radio for NPR in the morning and the kids use it as a Pandora radio in the afternoon. Please get CyanogenMod 9 stable on TouchPad. Please.

The iPad has games and sometimes has Netflix installed. I generally uninstall Netflix unless there's some compelling reason, like a road trip, to install it. But my wife has fairly well commandeered it as an e-book reader. We read e-books to kids before bed (eg, Nim's Island), which has proved a to be one way to condition the kids to understand the iPad is useful for something other than games. But, overall, if Netflix isn't installed, they mainly use it to play Pandora.

* They have access to two older, larger Dell laptops (dual and single core, 15" and 17", respectively) running Ubuntu, one in the spare room, one in serves as the main living room stereo, with Spotify piped to a big old Denon receiver. They occasionally use that one for music or the web; that's about it. There are probably a hundred games installed, they don't play them.

* We have a 27" iMac. They often use that for look-ups, occasionally a movie on the weekends, and occasionally some flash games. Whenever I see them playing flash games in the browser, I quietly ssh in and point that domain to localhost. The next time they visit, the browser errors out, and they move on. They also now use it to interact with the school's BlackBoard account (kill me now, please)

* There is a dual-monitor, two-CPU Dell in the garage, cobbled out of spare parts, running Ubuntu Studio, complete with condenser mic and keyboard.

We enforce a few rules

(1) No video or gaming during the day or school nights. The only exception is long trips.

(2) Sneaking off to your room with the iPad to watch video will result in the iPad being put in timeout, usually for several hours.

(3) Arguing over the iPad will similarly result in the iPad being put in timeout, even on trips.

Overall, the fantastically, overwhelmingly most common use of a computing device is music. The rest a 5-10% items. That is almost certainly due to the fact that we fight video and keep them otherwise engaged (Thing 1 is in gymnastics, Thing 2 is in piano, and rock climbs while Thing 1 is in gymnastics).

I will probably buy an iPad 3. I don't envision taking it to work. If CyanogenMod doesn't get stable on the TouchPad, I have considered buying two of the iPad 3s, but controlling what's on them.

So, again, I agree with the commenters who have observed from their own experience that young people are evaluating the utility of their tech just like the rest of us. If something's boring, or they are never exposed to it, they don't do it. Games and video are the lowest common denominator. It takes time to read. It takes time to do gymnastics and rock climbing and piano. We have to create a society where parents can and are encouraged, incentivized, to spend that time on their kids. Can Zynga get parents to spend more time with their kids?

(1) http://news.ycombinator.com/item?id=3590627 (2) http://news.ycombinator.com/item?id=3590416

I am pretty sure that is how it works, yes.

Sure, when I got my first Mac back in the day, all I wanted to play were video games. Because I was a kid and video games were cool.

Maybe it was the small selection of computer games available back then, but I quickly became disinterested in the games and wanted to start tinkering as soon as I knew it was possible.

Or perhaps we could simply persuade these children to think learning was cooler than video games. They are ours to mold at that age.

My high school lends MacBooks to all of its students throughout the school year. I will say that it is nice to be able to look something up at almost anytime. I have learned a lot from it, but it's nothing I couldn't have learned from using a shared desktop computer or my family's home computer. Still, for students who don't already have computers (the target market of the OLPC project), having an internet connection is really nice.

I will also say that the first year we were lent computers, everyone procrastinated a ridiculous amount. Games, everywhere, all the time. Very little work was being done. But now, three years later, people mostly just use the computers for work.

I don't think a laptop in itself is that valuable. Maybe I'm just jaded - computers are pretty fantastic. I'd say that by far the most valuable part of a laptop is being able to go on the internet: wikipedia, forums, google, it's all incredible.

Illiteracy would seem to be the first problem to solve in many of these areas, not a lack of laptops.

http://www.npr.org/templates/story/story.php?storyId=5251340

One library (card) per child might be a better goal.

However laptops are probably cheaper then building enough libraries.

My inlaws purchased a fancy laptop for my kids. It serves as a game and Youtube hub. Occasional Wikipedia lookup for school. I help them fiddle with a few kid-level programming apps, but other than that I have the existential feeling towards computers and kids: what's the point? I'd just a soon break out the Apple 2e from the garage and let them stare at the DOS prompt - they might discover its a computer that way and have to DO something that benefits them.

Not everyone is going to grow up to be a programmer. Viewing videos and playing games are important developmental activities. Yes programming a computer is a stunning form of creation for me and many others here but not everyone shares that view. For others, music, art, writing, teaching, and building give them that feeling of joy of creating.

Spot on. Not to imply computers are for programming only. I am perplexed at how to best make a computer/laptop useful to a kid. The issue I see are: too many options, too many passive activities, or too complex creative applications.

We may still not know the effects of having laptops in the classroom. Five years is little time to be measuring these things, specially considering that the ones who started using laptops have not reached majority of age. This is the same as saying that a new science curriculum does not work without having a real world test. We have to wait to see the results.

This should not be used as an excuse to implement one of this programs. Just saying that no results yet is different that net negative results or no new learning. Like someone in the comments in the article said, he became interested in computers because of his Commodore when he was little, which didn't reflect until he was of age.

P.S.: And the sight of kids on the streets, even the ones who are homeless or live in really precarious homes, using laptops and laughing is really a wonderful thing.

I at one point worked on the OLPC project fixing the laptops.

Lets hope that their tablets do better.

However, I doubt they will. The justification being that in my view the tablet is what the PDA tried to be fully realized. Well suited for work that one cannot sit down to do, but certainly not something I would choose in favor of sitting down to a desktop machine.

As the form factor of a device gets smaller, the quality of the sit-down user experience diminishes. This, if nothing else, is why I don't expect to see the desktop ever truly "Die" as a form factor, because it still makes the most economic (In all senses of the word.) sense for sit-down setups.

It's just that the number of people who want to sit down to a computer has dropped sharply.

Tablets are for consumption not creation. The key is the low bandwith input of "finger" vs. high bandwidth input of "keyboard".

I say they should be working on a touchscreen + mouse or trackpad + keyboard combo. Maximum input possibilities and ideal for creation and consumption.

Touch interfaces are just fine for creating. They only become limited when you're trying to emulate a different interface, like a precision pointing device, or half assed emulation of an interface like a keyboard, which relies on precise finger placement and a variety of non-visual feedback mechanisms.

The iPad has no practical multitouch input limits (I think it's limited to what int can hold), so it can track all 10 of your fingers at once, just as fast as a keyboard. Given the right interface design it has more input bandwidth potential than a keyboard, just not the current familiarity.

Give it time, tablets and multitouch are still in their infancy.

No, touch interfaces definitely have less bandwidth than a keyboard, and you can't make it all up with clever tech. Fingers are huge, relatively speaking, and for touch interfaces to make up for that fact they must get fewer bits out of the resulting input. The mathematics of signal processing require this. A stroke is much less definitive than a keystroke, or a mouse motion with click, etc. Moreover, you have the problem of interfering with output bandwidth as well as if you want to touch anything other than the very bottom of the screen you must obscure the screen with your hand while doing it for seconds at a time. You get less bandwidth both out and in.

What obscures this fact from your immediate recognition is that, unsurprisingly, touch interfaces are optimized to work under those circumstances, so you don't really "feel" the limitations, but you would if you really tried to push them the way a really good content creation application would.

For low-bandwidth applications they're fine, but they will never ever take over everywhere because they have weaknesses that make them unsuitable for high-precision, high-IO tasks.

I don't remember asserting they would take over everywhere, just that they can function fine for content creation. By content, I did not mean "stuff created with a keyboard" and I think my second sentence makes that perfectly clear. Touchscreens are clearly not perfect multi-purpose interfaces.

But I think you're ignoring the true significance of multitouch and other sensors and the additional dimensions of input that provides. A keyboard is nearly linear input. There's a lot of 2 finger chords and a smaller set of 3 finger keyboard chords, but for typing there's nothing really separating a 3 fingered man from a 10 fingered man except for input speed. Compare that to the capabilities of a 3 fingered man and a 10 fingered man playing the piano and you get the idea.

iOS is actually limited to eleven simultaneous touch events. We have ten fingers, typically, and there has been fair speculation about what the eleventh input method might be.

A cat?

tl;dr - giving an article a skim read it appears these devices wont allow facebook or youtube etc... so my point is probably mute..

Learning begins? Or we further Facebook's cause (along with other activities that are complete waste of time, gaming and streaming poor comedy). In Thailand at lot kids / teenagers / young adults have a netbook or laptop or a smartphone, if you are being a mr nosy you are guaranteed to see the little blue bar at the top of the screen and face palm, the wealth of information that is out there..

step 1: give every kid a laptop. step 2: gaming begins.

accessibility definitely plays a role but it's not the main impedance to education. you could say the same about access to basic pen+paper or regular books (libraries). this ain't no panacea - if a kid has the motivation to learn he will find a way. stupid and unmotivated children will approximately remain so regardless what tools or toys you give them (from experience of 1st world education)

this obviously within the bell curve.

depending on who sponsors this and what control they have over the content, a positive side effect of this social program (ironically) might be exposing the kids to the beauty of consumerism and opulent realities outside of their backward culture and spark the evolution out of it - in the sentiment of shanty towns sprinkled with satellite dishes

The metrics for determining the success of OLPC and similar programs should be long term, not short term.

I suspect that few of the HN'ers who started out typing BASIC from magazines into a VIC-20 suddenly had better grades because of it.

Instead, they were learning and developing interests which have had long term benefits - and that's what education should be about (at least in so far as one buys into Dewey).

It could be that those teachers who are spending thirty minutes getting students set up with their computers are spending that time because the computer makes it obvious which students are not prepared for class - i.e. it is easier to ignore a student who shows up without their notebook than one without a computer.

One TRS-80 or C64 VM per child.

Modern laptops, even simple ones like these, are too complex for a child to be interested in manipulating them, in my opinion. And I don't think the OLPC machines are powerful enough to do most enabling tasks that would help a student.

I've played with an OLPC laptop. The most interesting thing were the built-in programs to do stuff like measure distance using two OLPC laptops and sound. Not terribly accurate, I'm sure, but excellent for sparking discussion of physics.

Reply on news.ycombinator.com