Hacker News Reader

Comment by alexmingoia | Hacker News Reader

Learning to pass the test has failed us. Moving tests online won’t change that.

The future of education is not moving tests online. It’s no tests at all. The future of education is workshops.

Teachers as mentors, not lecturers and truancy officers. Exercises instead of lectures. Peers instead of tests. Projects instead of grades.

After completing a workshop, you have a project to show for it - not a score. After completing a workshop, you have an alumni network to draw on.

Workshops are self-directed. There’s a schedule to the exercises, but enough room for people to go at different paces. Students work in the open, and learn from each other. Students are encouraged to collaborate, instead of punished as cheaters. Students compete by seeing what’s possible.

Workshops are about learning to do great work, instead of learning to pass a test.

Workshops are the future of education.

View on HN

The point of a college degree is that you have a document, form an accredited university, that represents that you have completed a course of study in a satisfactory way. As an employer hiring a new grad, you know, at a minimum "This person showed up to class for four years. They are capable of following instructions, of learning new material, and doing so above a baseline level of competency."

What does an employer get in your scenario? Four years of 'projects' to sift through? How do they know what a good deliverable looks like for a CS 61A class at Southeastern Louisiana State University?

No grades works great at Yale Law, when you're dealing with a group of the most talented and motivated young people on the planet. That will not work in a mass education system, where the need to make the grade is the only thing getting half of the kids out of bed in the morning.

> What does an employer get in your scenario? Four years of 'projects' to sift through? How do they know what a good deliverable looks like for a CS 61A class at Southeastern Louisiana State University?

As someone who participates in interviewing undergraduate candidates, I prefer those that have projects to talk about over those who tout their GPA. Project work correlates very well with aptitude and initiative.

Projects require deep understanding, research, trial and error, and fully integrative thinking. There's no better way to learn and grow.

Yes but that’s because the current education system is geared toward GPA. If all of the students you interview have years worth or projects, how would you differentiate? Better yet, if they go through a university or trade school and they all complete the same projects, how do you know who is actually good?

I’ve run into this when interviewing or going through resumes of entry-level engineers. A lot of them will have a similar senior capstone project, so now I’m back to looking at GPA or some other signal.

“Yes yes you worked on CMU’s autonomous vehicles program. So did the other 10 candidates, but tell me what makes you truly unique? “ etc.

> Yes yes you worked on CMU’s autonomous vehicles program. So did the other 10 candidates, but tell me what makes you truly unique?

We might be looking for different things.

I'm just looking to hire self-driven folks that know how to program, problem solve, and communicate. I'm not looking for pedigree, elite status, or anything of that sort. In fact, I don't even care if candidates went to college. If they can demonstrate that they know what they're doing, then that's fine by me.

I predict cookie cutter projects won't be an issue. If a candidate can talk in depth about their project, the choices they made, the things they learned - that's good signal. It doesn't matter if everyone performs the same exercise, because not everyone will come away with the same experience.

That seems like a reasonable approach. Do you recruit directly or do candidates get vetted before they get to you? I get the sense that in many hiring flows, the candidate would have to go through some gatekeepers first. They may be the ones screening for pedigree, elite status, etc.

> I predict cookie cutter projects won't be an issue. If a candidate can talk in depth about their project, the choices they made, the things they learned - that's good signal.

I'd argue most capable people will come away with largely the same experience.

Like you, I have a bias toward projects and self-motivated individuals. What I'm saying here is that right now you filter toward that, because everybody has a good GPA, so it's not as strong of a signal. If you change the education system so that everybody has great project experience, you're going to swing back to some other metric to filter people out (like GPA/prestige of university/training program/etc.).

I don't think you can ever really escape it.

In my example, everybody has great project experience and a great GPA. So how do I decide? Well, I start looking at other things like maybe they're the captain of the volleyball team.

(speaking for university hires here at least).

Related question is, did you considered that most of them actually can be good for the job position in question? It is possible that majority of them would do reasonably well. In which case you are maybe wasting the time chasing for something special that you cant define.

And if they are not good, what it is exactly that they are missing? Then it makes sense to look for the specific thing most are missing. But it makes less sense to look for some special person without being able to define what special you look for and chasing random signals instead.

This is exactly the type of thing universities should not be for. We should create a separate institution that serves these needs without pretending to be a university. We literally need a trade school for people who just want a piece of paper that proves they can do what management tells them to. That isn't a university, it's just grades 13 through 16.

>As an employer hiring a new grad, you know, at a minimum "This person showed up to class for four years. They are capable of following instructions, of learning new material, and doing so above a baseline level of competency."

I was thinking about this point the other day. What made me question it is that everyone that goes to college went through 12 years of education that wasn't too dissimilar to that. But for whatever reason that is not considered.

A high school degree used to signal exactly that and enabled holders to get a “good job”. We mistook the map for the territory and insisted that everyone get a high school diploma, reasoning that this would enable everyone to get a good job. In doing so we destroyed the signal and the good job gatekeeper was moved to an undergrad degree. We are now in the process of doing the same thing again, and employers are starting to look for master’s degrees.

Only 6.4% of Americans got a high school diploma in 1900.

I don't know what GP wants, but projects can still be graded. And the project supervisor can assign a rough percentage breakdown of each project into subfields: 30% DLC, 40% programming, 30% technical writing. At the end of the program, you can collate these in a table to say things like 7 months spent on DLC across 5 projects with average grade of A-.

A university should not focus on 'what an employer wants'...

The lecture is still common in the college classroom, but there's been a lot of movement away from it over the last couple decades. For instance, I've for quite a few years now taught a data analysis class for economics majors, where the biggest part of the grade comes from a project rather than exams. Nonetheless, the lecture-exam approach will always be prominent. It's just a more efficient approach that works well enough for many purposes.

> Learning to pass the test has failed us.

I know the bigger point you're making, but it has dawned on me recently that life after school is always about passing the test. Every interview that I've faced has required me to study in order to pass the test to get in.

This sounds reasonable and it is also surprisingly similar to the past of education, e.g. apprenticeships in guilds.

Here is a very good RSA animation[0] on Sir Ken Robinson's talk on the same subject.

[0] https://www.youtube.com/watch?v=zDZFcDGpL4U

Wow that’s a great talk. Thanks for sharing!

I agree with you in many ways. Lots of learning can happen much more effectively in a project environment. I would say about 70% of a modern STEM degree can be based on projects.

But lectures have their place. A lot of times, you need information transfer from someone who knows a lot to someone who knows little. I am in favor of video lectures with active learning components to self-test yourself.

Also, subjects like Mathematics, require a lot of practice of, perhaps, boring questions. While some practice can and should be done collaboratively, a lot of it has to be done individually. There is no substitute for that.

That was always a better idea. So what makes you think now is the time the world will move from its entrenched tradition and over to something more rational?

Over time people will avoid spending their time and money on something that provides little value. It's already happening. The generation growing up now has learned (the hard way) that tests and grades have little value in the real world. That it's a game, one that they don't enjoy and they don't need to win.

COVID-19 has just accelerated this by moving schools online and laying them bare. Online there's no having lunch together, flirting, or getting dressed up. Online there's none of the things that made school more than a test. As the schools move online people see the emperor has no clothes.

Honestly, my experience of school was not just listening obediently as much as people make it to be here. And talking with my kids, the school was not like that to them - they miss the interactive parts of class. Now that homework is send by mail, bulk of it is like you describe. But the zoom classes are not like that and based on what my kids claim, it was not as dry in person.

Maybe there are completely dystopian schools where no communication is going on. Everything exists somewhere. But it is just not true that it is universal experience of school.

Can I say pair programming is a idea of the same sort. Code reviews take less time but vastly under-develop code awareness?

Reply on news.ycombinator.com