Allowing Innovation in Education

Recently, I came across criticism of coding bootcamps. As someone with a passion for improving education as well as both a former student and current employee of a bootcamp, I took the criticism quite personally. The main criticisms revolved around 4 main issues:

Aptitude: Unless someone is capable of teaching themselves they are not cut out to be developers. A person who has taught themselves is the only employee worth hiring.

Quality of graduates: the complaint was that most bootcamp grads are unqualified for the jobs they’ve been promised.

Lack of Regulation: coding bootcamps are unregulated and as such are likely to be a cash grab that takes money from students and then forgets them, repeating the process for the next batch.

Professional teaching: teaching must be left to certified instructors at accredited institutions. 

First and foremost, most would agree that education in the US is broken across the entire spectrum. There will always be those that are able to teach themselves and excel, but this isn’t really a solution to anything. K-12 education is not adequately preparing students, and college has become prohibitively expensive while providing less useful job skills. It’s clear that the industry is ripe for (please excuse the use of this word) disruption. To really understand how this needs to take place, we need to understand the nature of innovation and the current science around education. 

As far as education is concerned, the folks at Khan Academy have some great talks backed with data around learning science and how to maximize the speed at which individuals learn. You can view a great overview here about the tenets of efficient education. The main takeaways:

People are smart: categorizing people as gifted or slow in elementary school is an inaccurate and harmful practice, people are far more capable of learning fast and well given the right techniques. There are many adult learners that are incredibly capable once they’re shown how to learn effectively. More importantly, the elitist attitude of believing that most people can’t be developers/engineers/mathematicians/whatever is clearly false.

Methods: Flipping the classroom, allowing experimentation and expecting mastery, and having students teach each other is far more effective than the current lecture/assignment paradigm. And it’s even more effective than self learning, where the people get stuck or plateau much longer and get discouraged. One of the benefits of bootcamp style education is that it breaks bad habits and teaches people how to learn effectively for the rest of their lives.

The benefits of the ideal classroom espoused by Khan Academy has one goal: maximizing the rate at which each person learns. It’s no coincidence that this makes people far more capable than they would have been in traditional classrooms. Given that this style is the best way to learn, new models must emerge to implement it.

History suggests that significant innovation will not come from existing institutions. They are both too large and have little incentive to implement the drastic change that’s necessary. As long as university tuitions and student loan availability remains as high as current levels, there’s little incentive for the institutions to drastically change the current status quo for new models that are perhaps less profitable. Clearly it has to come from somewhere else.

In looking at the nature of innovation (examples here, and here) it’s clear that the odds against any newcomer are so great that any consistent success is a sign that the market is screaming for a solution. This is true for education as well. The market must decide what works, and as adoption of any new methods becomes more mainstream then regulation becomes necessary. In the early stages there is already so much skepticism and mistrust of newcomers that it implausible that any startup can survive through deceit or by trying to rip off students. This brings us to the requirements around regulation and accreditation of teachers. Both are incredibly important in mature industries to protect people from predatory behavior and to maintain standards, but incredibly damaging to any innovative education startup. We as a society have to allow experimentation and expect mastery; our education startups must be allowed to be incredibly agile and experimental when they are small and then held to high standards as they mature and enter the mainstream.

So to address the issues raised above:

Aptitude: Real data around education has shown that many of the myths around aptitude have far more to do with teaching correctly than raw intelligence. It’s disingenuous to say that most people can’t learn development when many in the industry managed to teach themselves and continue to do so. The real issue isn’t about aptitude but getting people to a certain point where they can continue on their own. The bootcamp model is awesome at this.

Quality of graduates: It’s incredibly hard to provide metrics on how bootcamp graduates compare to self learners or graduates of accredited programs. It seems the only thing that should matter is the market response. This is obviously program specific, so to say all bootcamps are bad is just as foolish as saying all bootcamps are good. What is fair to say is if a bootcamp leaves its students and hiring companies happy in the long term it is providing value.

Lack of Regulation: It’s incredibly difficult for any startup to survive. It is especially difficult for one to survive if it’s business model is to repeatedly trick students into handing over money. Especially with networks like twitter information spreads quickly. Startups don’t need heavy regulation, they live and die by their customers. We must allow them to be agile and hold them to high standards as they get larger.

Professional teaching: This is one of those things that would be great to have. Ideally, teachers would be trained for years in both learning science and their specialty, plus master their curriculum. Unfortunately, this is absolutely useless for technology education unless you want to teach theory or obsolete topics. The practical approach is to identify those with a talent for teaching and develop it in an agile way along with the curriculum. Any software developer knows that agile is more effective in the real world than waterfall, this is no different in education.

One of the most frustrating points raised was that anyone worthwhile can just teach themselves. A big issue with the ‘just teach yourself’ argument is that it implies a certain level of knowledge that comes from privilege. While true that there’s a limitless amount of knowledge available on the internet, knowing what to study or how to become a successful developer have a lot more to do with the circumstances of one’s environment than their ability. On a more personal note, I’ve spent some time teaching math and english to underprivileged children in India using the Khan Academy methods. Some had no idea what the internet was, but took to it quite quickly and were eventually performing above grade level even compared to their American counterparts. So the idea that if someone can’t teach themselves they don’t deserve to learn something is absurd, insulting, and elitist. People can be taught how to learn, taught quickly, and taught well.

I’m looking forward to proving it.


Now read this

Automating tasks with Capybara and Poltergeist (on Heroku)

The Problem # As the solo dev at a company that uses a lot of GUI tools to manage their work, I’ve been tasked with building tools to help automate their workflows. They must be easy to use since the tool may be used by contractors. They... Continue →