Hacker News new | past | comments | ask | show | jobs | submit login
Classroom Technology Doesn’t Make the Grade (bloomberg.com)
67 points by pseudolus 25 days ago | hide | past | web | favorite | 34 comments

Tech for testing is a good idea. You get the same adrenaline rush playing “race the teacher” type math games as in a live classroom test. You are timed, you know exactly how long you work on each problem, you can identify your strengths & weaknesses, & there's no need for a proctor to walk around the classroom. These are objectively good things.

I run a math site for 6-8 year olds & it’s an absolute delight to watch kids challenge each other to be the fastest at fun algebra & geometry puzzles. Several teachers have told me in person that kids who ace the puzzles on my site are ready for college. These are little kids in elementary & middle school! So it makes me very happy to have built something which is so powerful for reinforcing core concepts.

But then, some of the parents ask me how to get their kids to score higher. How to work some of these tests. Which youtube videos to watch. I tell them not to waste their time on tech.

If you want your kid to learn math, the best is one on one instruction. I sit with my kid for a few hours every week and we work problems on a whiteboard. Kids like markers & sometimes they will doodle with red & green markers while trying to think. This is incredibly fun to watch and it’s an awesome bonding exercise. I write a problem with a black marker & each kid has a color marker & they quickly start writing on different areas of the whiteboard and soon we have a solution. The $20 walmart whiteboard is an absolute godsend. We have a ton of markers & 3 whiteboards, one in basement & living room & upstairs, so whenever we feel like it, we can indulge in math. That’s the height of technology I’ve used.

Give your kids time & attention. That’s really all they want.

Education is getting its own spin on “You can see the computer revolution everywhere except the productivity numbers”.

There was so much research saying all the millions we were about to spend on tech in classrooms wasn’t going to work. Reduce childhood poverty and you’ll improve results. Focus on a knowledge based curriculum and you’ll improve results[1].

1. Why Knowledge Matters - ED Hirsch Jr

> Reduce childhood poverty and you’ll improve results.

Everyone knows this, but wealth redistribution is opposed by those who can prevent it from happening, i.e. those with wealth.

So we continue wasting taxpayer’s money and instead funnel that into some VC/PE/index fund owner’s pocket.

We could do a lot to reduce poverty without an extra dollar or redistribution. We could end the drug war, expunge non-violent criminal records, remove lots of occupational licensing requirements, and make it easier to claim benefits.

The goal would be to allow parents to spend more time with their children, and have more resources (money and knowledge) to share with their children.

I don’t see how this can be accomplished without some sort of wealth redistribution, by which I don’t strictly mean transferring dollars from one account to another, but rather forcing all businesses to provide more time off, thereby increasing labor costs and hence reducing income for capital owners.

This is more about using AI in the classroom. Often times family income is a major feature in determining low-performance. Simply not including this in predictive models isn't enough because you may find a proxy feature that is also highly correlated to family wealth.

Current research in edtech is now looking at the transferability of AI models. For example, if I target a specifically low-income / low-performing school and use some type of process to improve their scores, will that model transfer to another school (with hetero- or homogeneous students).

So the statement "Reduce childhood poverty and you’ll improve results." isn't directly referring to how wealth improves intelligence, but rather from a logistical regression perspective each "unit of wealth" is correlated to a higher performance.

There weren't a lot of vendors employing anti-salesmen to convince people not to buy stuff, or promoting research about how their products didn't work.

The sellers of snake-oil will always win over the denouncers of snake-oil, because very few denouncers of snake-oil are also willing to dedicate their lives to snake-oil.

As far as I've seen, educational tech products are mostly garbage. It might be that the worst players are sucking up the most money and attention via marketing and political shenanigans, but I honestly don't get the impression that anyone is even trying to genuinely execute well in this area. It basically feels like stuff made solely to tick boxes and not actually empower users.

There are several factors that I think contribute to this:

1. Education takes time. Therefore, evaluating whether an intervention actually improves education outcomes also takes time.

2. Decision makers both in companies and in schools buying their products are not end users, so their interaction with the product is limited.

3. Most learners hate the work it takes to actually learn something, so they'll prefer technology that helps them avoid that work.

In the end you get products that look good in a short demo and are well-received by students, but that don't really help in the long term.

> Most learners hate the work it takes to actually learn something

This is nowhere close to true. People love learning things and exploring new ideas, in a context where they have autonomy.

Essentially every human learns a tremendous amount about a wide range of topics and skills between age 5–20 years, with or without any explicit external instruction.

What they hate is being forced by someone else to do (what seems to them to be) pointless busywork.

Products which are marketed to schools as “educational technology” often seem especially pointless from a student perspective.

> This is nowhere close to true. People love learning things and exploring new ideas, in a context where they have autonomy.

There are many people who say they want to learn a foreign language or an instrument or some other skill and many free resources for self-study are available. Yet, most of those people either never start or give up halfway. Either they do not want what they say they want, or they're too lazy to put in the necessary work.

I'm a self-taught developer and learned a few languages on my own, but there's still a lot I wanted to learn but never did simply because I didn't feel like practicing and stopped.

That does not demonstrate that people “hate the work it takes to learn”. It only demonstrates that someone’s passively stated desire to learn some particular thing does not immediately grant them the intrinsic motivation, time, and external support to follow through.

Just because someone doesn’t spend every waking moment on deliberately practicing specific new skills does not mean that they hate putting effort into anything.

Most “free resources for self-study” are crap, and only work for people with unusual amounts of motivation and focus. This doesn’t mean that everyone else is just lazy. People lead busy lives and have many distractions competing for their attention.

If you just said “many people claim they want to learn various specific skills and subjects but then never make time for it”, that would be an uncontroversial and easily factually supported claim.

* * *

Motivation is pretty tricky. It would be beneficial to everyone if society spent more effort on teaching meta-skills like how to break large goals into small manageable chunks, how to manage time and focus, how to seek out resources, how to evaluate progress, etc., and more effort on offering people meaningful support for achieving their own goals.

From what I have seen most “educational software”, testing-focused pedagogy, de-professionalization of teachers, etc. are harming rather than helping.

I'm building something to change this.

I'm addressing 1 by building time into the core of the product. I have a first-class 'timeline' model, and they span multiple years.

I'm addressing 2 by selling direct to consumers. This will be hard but I agree that 2 is a big problem and so I must start by avoiding it.

3 I'm not really solving. I'm just going to initially target highly motivated people where there's a clear reward for their hard work. Eg. tech students and med students.

> I honestly don't get the impression that anyone is even trying to genuinely execute well in this area. It basically feels like stuff made solely to tick boxes and not actually empower users.

I am.

I think I have a decent chance at success because I'm building it for myself first, to improve my own knowledge and skills. I'm a motivated learner and I care about having something that respects the hard work and time it takes to educate oneself; most ed-tech feels like fairy-floss instead of porridge.

Whether or not it ever succeeds commercially, it hope to get it to a point where it's a properly good tool to accompany a student going through higher-education and continued professional learning.

Aside from the political nature of the procurement process that gives the advantage to connected indivduals, educational technology is basically a confluence of the worst kinds of conditions to build a business: low-budget clients (school districts), high-maintenance users (kids and their parents), and burdensome (in my opinion, legitimate and necessary) regulatory concerns around data integrity and protection.

I fully believe you could get a first-rate education with only a chalk board and a tree for shade. Similarly, soldiers and body builders in the Soviet Union did their training with a pull up bar, a bench, and a few free weights. You don’t need much equipment for anything if your technique is good and the same applies to education.

For as long as we humans have existed, we've sought for better and more effective tools, and now you're saying that's overrated?

Not at all. But we also have a bad habit of conflating flashy new tools with effective ones.

I'd also add to this that these technologies can be disruptive in another tangential fashion. It's not like a school can ever reach the point of being 100% technologically integrated. You can always have more tools, more software, and more technology in general.

And so if at some point you do not step back and say, "You know, I think our problem might be that our current trajectory is fundamentally failing to improve the educational process.", then you can effectively spend an infinite amount of time, energy, and money pursuing a task that is not, and will not, ever provide the results that you're seeking - yet there you are convincing yourself that all you need is a bit more money, a bit more time, or the hottest new tech.

And the view will always be controversial because there is an immense amount of money to be made in providing these tools and that money will fight back. It's also quite intuitive that technology should be able to provide substantial improvements in education, and so those driven by their intuition will also fight back. Unfortunately, something being intuitive and something being correct are far from the same thing.

As humans we shape our tools and in turn our tools shape us. Not all tools are good for us. There is a growing backlash against technology in the classroom and it’s being spearheaded by wealthy parents. They recognize that tablets are consumption devices optimized for engagement and that they tend to be very addictive to young children, a bad mix if your goal is to raise your child to be a leader rather than a consumer.

No, only that these aren't better or more effective tools.

the search is good, but sometimes you don’t find anything.

"But until these tools offer clearer and more widespread benefits for students, they should be evaluated with serious skepticism."

This is the biggest takeaway I saw from this article. Specifically because measuring improvement in education is difficult. The most used metric for this is through the use of something known as a "learning gain". Typically, this is measured using a pre-test/post-test on a topic, with the intervention method in between. However, because an experiment needs to be strongly maintained, these experiments are often done on smaller scales, like for two to four weeks. Thus, the learning gains are only measured for a single topic. Semester-long or longitudinal studies are more difficult simply because it is harder to measure and control improvement over time while ensuring each student is given a chance to learn the material.

It’s the same mistake made by both tech companies and grandmas — that “engagement” == “value”. People who spend more time on sites, and kids who spend more time on iPads, does not make them experts in tech; just their ability to consume.

This article leans heavily on NAEP tests which the US has never placed at the top on. Yet we still have the highest number of the world's best institutions of higher education.

In education research, if the NAEP is the only metric you're using, your findings are probably skewed from the get-go.

I find it a bit circumspect that the article specifically mentions that a lack of adequate teacher training with EdTech tools is certainly to blame for its limited impact, but mentions it only once in passing and instead concludes that hitting the brakes is the solution.

Our elementary school provides iPads in all grades, taking them home starting in 3rd. I was horrified to see my 5th grader zooming into math worksheets and using her finger to work out the problems between the text. We put a stop to that and had her work it on paper, then submitting a photo.

Now in 8th grade, at least it is more paper based, but they are still provided laptops. She told me one of the teachers pre-records the lectures, and they watch those in class, instead of live instruction. I suppose the teacher thinks she can record it once and use it for several years.

Now I know why IQ peaked in the 90’s and has been dropping after the introduction of the smart phone.

What’s being used in your child’s 8th grade class is the flipped classroom strategy [1], as already mentioned. In my own experience through two graduate degrees, I rarely saw the benefit of attending a live class. I now come to find, many years later, that students’ class attendance in some graduate programs is at an all time low [2], and seems not to make a difference [3].

I encourage my 5th grade daughter to do both. She’ll workout problems with pen and paper or on a white board, and complement her studies with online tools [4,5,6]. Based on this regimen with a sample of one, she’s done very well. ;-)

[1] https://en.wikipedia.org/wiki/Flipped_classroom?wprov=sfti1 [2] https://www.statnews.com/2018/08/14/medical-students-skippin... [3] https://www.usnews.com/news/education-news/articles/2018-06-... [4] https://beastacademy.com/ [5] https://aopsacademy.org/ [6] https://www.curriculumassociates.com/products/i-ready

> flipped classroom

>> She told me one of the teachers pre-records the lectures, and they watch those in class, instead of live instruction.

That's not flipped classroom, it's some weird cargo-cult hybrid combining the drawbacks of pre-recorded lectures with the drawbacks of in-classroom lectures.

The whole point of flipped classroom is to move mostly non-interactive stuff like lectures out of the classroom by assigning them as homework, while moving exercises that were previously considered homework into the classroom where a teacher is available to answer questions.

I hope OP's daughter at least gets a copy of the video she can re-watch later.

I am sorry I missed that essential detail in your sentence. I agree, that’s not the flipped classroom strategy. You’re right to be uneasy about that teacher’s approach which does sound like a class on autoteacher for a few years. I had a teacher in 10th grade that would ask us to read any book we had, while he put his head on his desk and “meditated.”

Why is it horrifying that a 5th grader solves math problems by drawing on an iPad instead of drawing on paper? Why is paper always better than tech? Would you be horrified to have her write an essay or report on a laptop instead of handwriting it? If not, what is the difference?

Having students watch lectures at home and focusing on individual problem solving during class time is an educational approach called the “flipped classroom” and is supported as effective by research. I wouldn’t worry too much about that methodology.

On a different note, the classroom technology that seems really important as the parent of young kids is the parent communication tools. Comparing notes with my parents, we both agree the communication about expectations, curriculum, and homework is really enhanced by technology

This is basically how modern homeschooling works. I'd argue over 90% of traditional in-class time is a waste after kids are literate.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact