Finishing school is signaling that you are person that shows up in the morning for classes and puts up years of his life to do that. By any means I am hiring developer who is signaling that he is reliable and comes to work every day vs someone who is inconsistent.
I think the problem here is a bias due to the fact that the employment after earning their computer science degree is for many people not about the science part of it but rather about fairly mundane programming and general office work.
Good luck arguing that chemistry, biology, mechanical engineering, medicine and many other fields that require specific tools and environments in order to actually study not to mention make any progress in the field are all about signaling.
Emphasis mine.
No one cares if medical students avail themselves of digital resources to help them through medical school. Good for them.
Show me how many practicing doctors passed the boards and made it through residency after ONLY using Youtube channels and flash cards to self-educate.
Some people rely on the 'free' resources more than others. The real question is where is the cutoff percentage at. Say someone got 95% of what they know only from free resources and passed boards swimmingly. That sure don't look good on the med-school.
Say 95% of the entire class got 95% of the learning through Youtube/Anki and 95% of them passed boards. At such a percentage, med-school is all but useless to the general public that they serve. May as well get rid of them.
Granted, I don't think it's anywhere near that kind of level of dereliction that the med-schools are at (Cadaver Lab is an obvious counterpoint). But, where is the cut-off point for the schools and society? It's not 5% of the material being learned outside of them, that's fine I think. But if 95% is 'learned' outside the lectures, then yeah, that's a real bad sign.
It's a complicated question and the answer will likely be more complex and will evolve from class to class and year to year.
Have you a real, physical example of someone who "got 95% of what they know only from free resources and passed boards swimmingly"?
Here in the UK, a large proportion of healthcare is delivered by healthcare professionals other than doctors. If I go to my GP (family doctor) with a minor ailment, I'm likely to be treated by a Nurse Practitioner, who may have a Master's degree in nursing or may have never attended college at all. If I have a minor surgery, the surgery might be performed by a Surgical Care Practitioner working under the supervision of a consultant surgeon.
Lambda School have conclusively shown that it doesn't take four years to make someone into an employable software developer. How many other job skills could be taught through a short bootcamp programme, intensive vocational training or on-the-job training?
That’s not accurate. Law school requires any degree followed by three years of a law degree. Now that’s pure signalling. Every other Anglophone country bar Canada has undergraduate law degrees instead of requiring what amount to two undergraduate degrees.
Medical degrees in the US are similar in that you complete an undergraduate degree before you can get into medical school and get a MD or DO. Some countries have the MBBS degree which is a bachelor level degree that can be started right after the equivalent of high school.
A three year degree with plentiful coursework and no research component is not exactly a central example of a doctorate. M.D.s, D.D.S.s and J.D.s are called professional doctorates in the US and a second-entry bachelor’s degree in Canada.
> Medical degrees in the US are similar in that you complete an undergraduate degree before you can get into medical school and get a MD or DO. Some countries have the MBBS degree which is a bachelor level degree that can be started right after the equivalent of high school.
Some countries are everywhere apart from the US, Canada and former US colonies like the Philippines. Undergraduate medical education is the norm everywhere though postgraduate programmes are extending across the globe.
There is also other side of coin, doctors who are really experienced in some area are going to be expensive. There is also a lot of people who you can treat by googling stuff. Of course they should not take pills without second hand opinion, but not everyone needs brain surgery. If you get viral infection or bacterial infection you have to send stuff to lab, and lab returns results in readable way, so even GP (general practitioner) is not needed. Once I even got my blood tests and "GP" (my first contact doctor since I am not in UK) was like: "yeah I dunno, I usually get older folks with those specific problems so I cannot really help you". Boy that was nice, because you can go to some ass who thinks he has to know everything and that would be annoying.
Be humble, even as developer you don't get to know everything. Don't think doctors, mechanical engineers know all.
I've worked with people with previous job experience that have gone through 10-week coding programs. And it shows. They might've been taught a framework or a language, but their computer science skills aren't nearly as developed. Some things take years to click, and growth occurs from years of writing bad code.
Were they useless? Were they worse than useless? Or were they just not as good as a more experienced developer?
I'm not arguing that college has no value, but that it's bad value for money. If someone can become a useful-but-flawed developer after a short bootcamp, surely it's better for them to learn on the job while earning a living rather than mortgage their future on four years of education.
If college were free to the student and cheap for society, sure, send everyone and don't worry about it. That's not the case though - an entire generation have been saddled with vast, unmanageable levels of student debt. We need to be asking serious questions about how much and what kind of education is really necessary to produce skilled workers.
You don't need college to write years of bad code. Internships or hobby projects do fine.
Depends on my ailment. Most of the time people go to the doctor they are told to get some rest and maybe prescribed something. I don't think that needs 8 years of medical school. And doctor's don't even necessarily do a good job at prescriptions either, since they tend to over-prescribe in the US
What takes 8 years of medical school is knowing when not to tell the patient that. When telling the patient to go home and rest could result in severe illness or death.
Nobody in charge is doing the "extra training increases costs by X which decreases availability by X which causes X damage vs lack of training causes X' damage" calculation -- everybody just argues for more training every time there's a problem, with the inevitable result of our stupendously expensive outlier of a medical system with middling measured performance and poor, ever-worsening accessibility.
That's not a fair comparison. A better question would be: "Would you go to a Doctor that spent 9 years learning on the job or 9 years in school?"
Besides I'd assume with this setup the future doctor wouldn't start with brain surgery. Maybe stitch up cuts and set broken bones, working up into more complicated scenarios.
I mean, I've always done my own minor doctoring anyway. And people did for a long time, it's only recently a long stint in college was required.
Granted this approach had variable levels of success, but the idea you have to have formal schooling to fix up any health problems I'm not sure I buy.
The article is about the fact that school does not show that "developer ... is signaling that he is reliable and comes to work every day"
The whole point of the article is there is zero correlation between people who've graduated from college and people who haven't in term of how well they do on the job.
It sounds counter intuitive. There's like 3 situations
1. College trains you more than non-college
2. College doesn't train you but shows you're willing to stick things out more than people who didn't go
3. College does nothing what-so-ever (no difference in job performance from hiring people who did or didn't go to college)
The author of the article is claiming 100 years of research has shown it's #3. People will let you in the door because you have the paper (diploma) but they are fooling themselves that that paper has any meaning relative to hiring people without that paper
Here's another interview with the same author
http://www.econtalk.org/bryan-caplan-on-the-case-against-edu...
again, I am not agreeing nor disagreeing with the article. I'm only clarifying what it's trying to say
Let me add though, the author is claiming this is true in aggregate. Not for your personal anecdote.
... among the population who have been hired to do the job. This is conditioning on the collider. The people without a degree who got in are a highly selected sample compared to all people without degrees. Anyone who used this to argue that a degree wasn’t a strong useful signal Wouk be making the same mistake as those trying to get rid of the GRE in graduate admissions because in the population admitted the GRE doesn’t predict anything. If it did that would show under or overweighting ofbthe signal it sends. Zero correlation shows it has been given appropriate weight.
Reread the title of the article.
The question is whether an individual would benefit from college even if everyone's college was enrollment/graduation status was kept secret from everyone, even classmates.
This is not what the article says. It's saying employers hire graduates because they're better employees, but college mostly didn't make them better. This is closer to your option 2. (We are talking about the OP, the Caplan editorial, right?)
Caplan is saying college is mostly an expensive arms race. It has to be expensive (in students' time and attrition through boringness/difficulty, if nothing else) to be an effective signal. Subsidies in recent decades have made it even more expensive.
Not sure about others but I tried my damnedest not to take morning classes unless I absolutely had to -- most semesters I had nothing before 11am or noon. The flexible scheduling was one of my favorite things about college. I did take one 8AM class and thinking back I have no idea how I didn't completely bomb it -- I showed up pretty sparsely after the first week or two.
Doesn't everyone complain about how technical interviews only measure how well the candidate is prepared for technical interviews? With that in mind, I'd argue that the willingness to prepare for a technical interview has become just another signal. Also you go against your argument just after:
> My first company that hired me out of college got a much better deal waiting for me to complete my senior year. Yes, each year isn't worth the same. This makes sense. Senior year was my capstone -- building valuable communication skills in a team environment.
Firms can not measure soft skills accurately, so this returns right back into the value of the college degree as a signal. You get points for attending a school which provided a teamwork-oriented final course. It's your final grade that matters, not what you learned in that course. You might have developed better skills than the rest of the cohort, but it's meaningless if you didn't get a better grade then them. Everyone would rightfully assume that the others have better developed skills than you, because that's what the current system is able to measure.
One final unrelated point:
> Would you pay 50% the price for a half-finished version of Microsoft Word? Probably not. It'd be far less useful of a piece of software.
The utility of added features is diminishing. It'd be worth a lot more than half, or Microsoft would just keep on piling features and raise the price quadratically (or with some power above one).
Many do. I used to believe that was because they had never sat on the other side of the table and seen how many people were just trying to bluff their way through without the ability to do anything.
Now, I am much more bitterly cynical.
The 80-20 rule applies to software: 20% of the feature set is used 80% of the time. Mid-way through development, it's likely the 20% may not be complete. It was a poor analogy.
License and bar exams are more serious than people in technology are willing and ready to admit. They serve the role of quizzes because the "passing score" requires a baseline of knowledge that the quizzes attempt to establish. This isn't a matter of signaling so much as the effectiveness of the testing process. The fact that software companies cannot rely on these factors is a poor reflection of the industry.
Practically speaking, there's nothing stopping the JS foundation from developing a curriculum, standard test and credential.
The other issue is that there are layers of other skills that also evolve every few years.
So to create a "bar exam" for a front-end web engineer, for example, you would need a completely different version every two years or so.
But then probably a bigger problem is that software engineering requires certain types of problem solving and to get a realistic idea of problem solving abilities of a person at large enough scale problems you would need a huge investment into the testing infrastructure and varied content. You would need sophisticated programs to test the applicants programs. And people to review code. And those tests would quickly go out of date.
Then the other thing is that there is so much variation in the types of languages or frameworks or tools used, it is questionable whether it is meaningful to test on some lowest-common-denominator set of tools at all, if you could find enough agreement on such a thing (e.g. Angular vs React, React vs Vue).
A CS bar exam should just assess fundamentals -- if you have a good CS base, and problem-solving skills, you will adapt with the times as technology changes. Then, it wouldn't need to be updated to include new ES6 functionality. Besides, someone who does back-end Java doesn't need to concern themselves with language change to JavaScript. That's another reason why such an exam should be generic.
I'm honestly surprised nobody has tried to make that a thing. It'd save companies so much in time and effort if they could just say "yup, this person has the credential".
Most of them were vendor specific and so sucked.
Others were too low-level and so were meaningless, or never had any buy-in from anyone. They sucked.
Finally, can you image the same crowd that complains about technical interviews responding to a bar exam? "I haven't ever had to write a binary tree in my 30-years of real-world experience! Why would I have to do it for your test?!" That would suck.
you don't get those in any school. your school's signaling secured that.
Does going to a top 10 law school matter more than graduating from a Top 10 CS program?
Many other industries tend to recruit only out of top universities, and getting into them if you didn't go to one is basically impossible. Management consulting, investment banking, hedge funds, philanthropy, many managerial government positions. The way you get into these is through going to a school that places lots of people into those industries, and then work the network.
(at least the bimodal bit and ivy->faa[n]g route)
I have both law and software engineering degrees and I can tell you without a doubt that yes, the school you graduate from matters much more in law.
"Software engineering" doesn't have much to do with either software (it's mostly generic project management) or engineering (ask an engineer: vector calculus! statics, dynamics and thermodynamics!)
Signaling doesn't buy you much. At least in engineering, because a technical interview will sort out whether or not you actually demonstrate basic programming skills very quick.
Software Engineering is a notable exception, most other fields don't quiz as part of the job interview process: soft skills and credentials or state licenses or bar exams are relied upon more heavily. Maybe signaling matters more in these fields? Again, signaling is no measure of real skill.
Obviously, like most people I'm upset about the pay-to-play scheme but in reality very, very few people choose that option. 99.9%+ of applications are not pay-to-play, so we should still hold a good deal of faith in our higher education system. It is imperfect and has many issues, but this article title is broadly incorrect about the purpose of higher education today.
edit:
> Researchers consistently find that most of education’s payoff comes from graduation, from crossing the academic finish line. The last year of high school is worth more than the first three; the last year of college is worth more than double the first three. This is hard to explain if employers are paying for acquired skills; do schools really wait until senior year to impart useful training?
My first company that hired me out of college got a much better deal waiting for me to complete my senior year. Yes, each year isn't worth the same. This makes sense. Senior year was my capstone -- building valuable communication skills in a team environment. Year 1? I was finishing the basics of CS 101. Of course the last year is worth more than the first.
Would you pay 50% the price for a half-finished version of Microsoft Word? Probably not. It'd be far less useful of a piece of software.