Ask HN: Is it worth paying for a Coursera course?

138 points by eecks ↗ HN
There are lots of free courses online but some offer the option of paying and receiving a certificate on completion.

Does anyone here have one of those certs? Do they hold any value? Is it worth paying for the cert when you can still do the full course for free?

114 comments

[ 3.5 ms ] story [ 62.1 ms ] thread
From answers given by some CS professors on Quora, just a cert from a MOOC means nothing to them if you are applying for PhD. What they really care is if you have mastered the techniques you got from learning the MOOC.
Is the cert a step to proving you have mastered the techniques?
Perhaps doing something with the knowledge gained from the mooc. Like contributing to open source projects, side projects, etc.
For many Coursera courses, absolutely not. The cert just says that you're good at knowing where to look for the answers when you take the quizzes/tests. I'm sure there are some courses where this is less true, but most I've taken are simple "what do the provided notes or videos say the answer should be?"
So it is pretty much the same as "regular" college courses?
Except you can retake the exams until you pass.
I'm not sure Coursera is really aimed at already academically strong students trying to find a professor to accept them as a PhD student though.
Disclaimer: I've paid for many courses but never paid for a specialiazation.

I personally don't care about the effect of the certificate. I paid simply because what instructor taught me worth more than it cost.

What courses did you do?
Most of them are about Data Science / Machine Learning and some math stuff
Same here. For example, I was not planning to pay for Andrew Ng's class. But at the end, I've learned a lot and enjoyed his style of teaching, and figured that by paying I was helping support his mooc course for others to enjoy as well.
I personally paid one course and one specialization courses.

There are the reasons why I paid:

1. The course is awesome and I want to support the authors.

2. The course I paid is new learning area I never use professionally during job but I wish to use at my job (or new job). Instead of say "I interested in XYZ", I said "I interested in XYZ so I took online course to learn" at my interview. This is to partially solve chicken and egg problem.

3. For serious courses, I pay so I can have pressure and motivation to complete it.

If you've reached that stage yet, have you found prospective employers appreciate the (paid-for) certificate?

Over taking the course and saying "I'm interested in X, so I took the course for free, and consequently have no proof to show you".

I have some Machine Learning certificates from Coursera, and at the gig I'm working some really entry level Machine Learning was required.

I do believe it helped when I told them about the courses I took, showed them some code I had backed up online, at least it let them know I wasn't just completely bluffing.

In me experience, employers does not care whether it is free or paid, they care more about your passion to learn.
Yes, though the paid certs aren't much better than the unpaid.

Source: hiring manager for Data Scientist/Engineers.

I find lots of PhDs who can't code and lots of software devs who can't build a regression model. If you are a software developer who has done the Coursera ML course and/or a Kaggle then your resume gets read.

Surely the next question in the interview would be to ask a bit about the course and what was taught in it. No one asks to see my University qualifications in interview, an instead ask fairly irrelevant algorithmic questions (I am a database / systems engineer).
It depends on the role. For our team, IMO the difference in attention is epsilon, since we already see so many applicants who have a minimum required exposure to ML.

Source: not the hiring manager but involved in reviewing candidates for a DS team.

If you don't mind which courses did you pay for? Especially, which ones did you pay because they were so high quality?
If you don't mind which courses did you pay for? Especially, which ones did you pay because they were so high quality?
For some people, but not others, paying for the course helps motivate the person toward completion.

For some people, but not others, the certificate holds value independent of the learning experience. That also can motivate the learner to completion or increase esteem for an applicant to an academic or commercial solicitation.

I have a bunch of the free unverified certificates from a few years ago. I found the certification process motivating and a useful time management tool.

Good luck.

I took four or five Coursera courses in 2012-2013 while trying to get into the industry. I listed the completed courses on my resume. Employers seemed mildly intrigued but never made a big deal about them, and no one ever asked to see certificates (back then they offered free certificates of completion if you completed the course with a sufficient grade--there was no webcam of typing verification back then, and I think the certificate was just a pdf you downloaded or something).

My experience may not generalize, but I'm not sure I'd worry about it.

Note: I also had an undergrad degree (math and philosophy, and graduate degree in philosophy, but no school CS experience. No idea how this changes if you don't have any degree).

For the 3rd party (ones that you want to show off the achievement), I don't think there is any difference between "Passed course on Coursera" vs "Passed course on Coursera and paid for certificate".
Has anyone completed any business courses / specialisations on Coursera? Are they worth it for a developer trying to make the move into a business-oriented role? And, do you think, that certificates matter for securing such roles in companies?
I took a gamification course from Wharton that was quite good and I've looked at some others. Sure, if you're looking to move to more of a business role, courses in marketing, finance, etc. would probably be useful. I doubt if the certificates matter though and the courses won't have the sort of group projects, case study discussions, etc. that you'd get from an in-person class.
What kind of business role do you want? If it's sales pickup books like Spin Selling, Launch (by Jeff Walker), and Predictable Profits Playbook.

If it's more general finance you want to learn coursera should be fine.

If it's leadership qualities, start with the book How Google Works.

I finished the R course offered from Johns Hopkins when it was first offered a few years back. I was doing my MS at the time and it gave me an edge over students who were just beginning the fundamentals of R. I could do a more in-depth analysis for my project, so it certainly helped me. It wasn't paid. Would I have paid for it - for what I learned and how R is a big part of the work I do, I want to say 'yes'. But I think you need to have a clear understanding of what outcomes you expect before committing to a paid course.
hello !

I have completed two business courses on Coursera, namely:

1. The Global Business of Sports

2. Foundations Of Business Strategy.

Both these courses I took more out of curiosity, rather than anything else.

Both these courses were good and helped me understand a lot of perspectives, though I have spent 16 years, in hardcore sales and distributions across 7 different industries.

I believe the certificates do matter, in the sense that, you started something and finished with a minimum qualifier.

If you are a developer and managed to finish business courses and have something to prove for it, thats says a a lot about your capability and attitude.

cheers !

I started the Intro to financial accounting today - which prompted me to ask this question.
I have several from when they were free, and none from when they started to charge for them. I don't think it means anything, really. You can print your own score screen and show interviewers and it would mean as much as a coursera-generated pdf..
I agree with most of the other answers here. If paying provides personal incentive or you want to "give back" (although I suspect that most of the money goes to Coursera and its VCs rather than the creator of the content), sure, go ahead. However, outside of edge cases such as needing a verified certificate for a continuing education requirement or something along those lines, it's hard to imagine the certificate itself having a lot of value.

(Which is one of the challenges for MOOCs. On the one hand, most adults aren't going to pay a lot of money just for a learning experience. On the other hand, there doesn't seem to be a lot of value in the MOOC certification process. I suppose that a specialization from a good school might be worth something in the absence of other relevant background that you could point to, but I have to believe it's marginal.)

I think they have some financial aid for some people. That might be worth looking into.
A follow up question to the OP's:

Which courses are worth taking?

I've taken Andrew Ng's class and thought it was fabulous despite having extensive prior ML knowledge. I'm currently taking a negotiation class by Yale university which is very valuable do far and already has changed how I think about deal making. My wife just finished a script writing class and also got great value out of it despite being an accomplished writer.
I'm just finishing up Andrew Ng's Machine Learning course and would definitely say it's worth it (depending on how much, if anything, you already know about ML).

I'm also about halfway through the Johns Hopkins Data Science specialization, and I've found all of those courses to be worthwhile. But then again, I'd never programmed in R before and that's something I very pointedly wanted to pick up. So far, this series has been invaluable in terms of learning R (much like I had to learn a fair amount of Octave for Andrew Ng's class).

I paid* for and passed the cyber-security specialization from the University of Maryland. Overall I think it was worth it since I learned a lot and met some interesting people. Half of the courses were substantially less useful than the other half (Hardware Security / Usable Security were not as informative as Software Security / Cryptography). I added the specialization to my LinkedIn profile & Resume but I've never had anyone ask about it.

Coincidentally at my old job I was the cyber-security expert for my domain (HW), which only meant that I would be assigned to security relevant programs if we got any (we didn't), and they paid to send me to Defcon. They didn't even know I was taking the specialization, just that I like to talk about security.

*I say paid because I just got a refund since apparently there are some issues with migrating courses to a new format and not offering enough of them this year.

Do the certs hold any value ? compared to a diploma or real work experience, the answer is no.

Can it be a plus on a CV the anwser is yes as it show you're willing to learn new techs/ skills continuously.

Coursera has great courses, no question, but clearly their certs offer very little value compared to official diplomas.

Coursera's assessment methodology started off bad, and when they moved to the new platform it became even worse. It's now basically impossible to write meaningful or good examinations using their platform.

Previously, you could do novel things like limit the number of times students could re-take exams and assign different point values to different questions. No longer!

All a Coursera completion tells me is that you were persistent enough to go back and re-take the (multiple choice) exams enough times to get a passing grade.

This is honestly true.

I take quite a few Coursera courses and I have had a score of '100' in all of them. And this is almost exclusively because even if I fail I can just see which question I failed, go back to the lecture, study that section again, and try again until I get the 100.

On the other hand, what you're saying is that the tests showed you what areas you were weak in, and prompted you to go back and study until your knowledge improved in those areas. Put that way, it sounds more like a feature than a bug, no?
You don't need to do that, though.

There are only N choices per question, so when you get it wrong you just go back and pick another one. When you see that you got it wrong, you go back and pick another one. Eventually, you'll know the right answer.

You don't actually need to watch the videos to do this, and indeed, some students just go to the test and brute force their way through the test to get 100%.

I guess my reaction to that is "So what?" So long as the grade has minimal real world meaning, why should I care that someone is "cheating" to bulk up their directory of Coursera PDF certificates?

I agree that gaming tests is something that Coursera should presumably care about because their business model presumably depends on certificates meaning something in terms of people getting jobs etc. someday. But, as things currently stand, it's hard for me to get excited about people fooling themselves to rack up virtual MOOC karma.

If the grade has no real world meaning, then why pay for it? That was the OPs question.
Some of this turned into a broader discussion.

But, to the OP's question, no--I don't think there's value in paying beyond edge cases (e.g. proof of continuing professional education) and any personal gamification benefits.

Yeah, if you can actually pass an entire course as a non-comprehending algorithm, it's not much of a "certification".
Of course, one might argue that, if you're using testing as a pedagogical technique as opposed to a certification mechanism, that's exactly how it ought to work.
Exactly. If we go from the assumption that the whole point of lectures and courses is to acquire knowledge, doesn't it make sense to make sure knowledge is indeed acquired.

There's that saying that amateurs practice until they get it right and pros practice until they can't get it wrong.

A perfect score in the current configuration tells you that the students didn't just persist through the course, they went back and filled the gaps and made sure they get it right even if they got it wrong at first.

But is that long-term knowledge?

In college I took a class and, after the first session, realized the professor was utter shit[1]. I did not attend a single subsequent class or complete any reading/assignments. After cramming for the midterm for 10 hours, I earned an 80% and dropped the class to prove a point[2].

Could not tell you a single thing that was part of that class, but I got a lot of questions right.

[1]: He was hired because the college wanted his wife, so they hired them both. Terrible hiring practice. [2]: Yes, I was a pain in the ass at that age.

Preaching to the choir, tyre: I have a catastrophic college track record, 9 years instead of 5, almost expelled, extremely poor grades, missed exams, pissed off some teachers because I didn't think their course was worth the time (because I knew more about the subject than they did and they weren't okay with that even though I shut my mouth unless they said something wrong and I would suggest something correct very discreetly, privately, and as a question not even a statement, giving them the opportunity to take credit for the embedded answer. Yet they didn't like it).. Those very same teachers would entrust me with the class to answer questions or assist students when I was there, and then give me poor grades.

It's hurting me right now because a recruiter would look at my grades and think I'm stupid, so you do have a point. I was looking at it from the personal quest to quench the thirst for knowledge, not how it looked from the outside.

> All a Coursera completion tells me is that you were persistent enough to go back and re-take the (multiple choice) exams enough times to get a passing grade.

Doesn't it indicate that a student is willing to learn?

Learn the material or learn the test?
Bit of both.

My understanding is that if you hear something multiple time, it will stick with you.

not all coursera courses use multiple choice.
Just couple days ago I decided to give 5 weeks android course on edx a shot. I'm in week 5, spend like 3-4hrs to complete 1-4 weeks work (including time spent watching videos), wrote probably no more than 30-50 lines of code mostly something simple as "smth++". I feel like the quality went down dramatically compared to first courses they had, now its all about the money...
Andecdotally, there does seem to be an increase in the number of courses on these platforms. I'm not sure that's wholly negative given that the data has shown pretty consistently that with each week more learners stop engaging. Which suggests that smaller chunking might be more effective.

Of course, it also means that there are more opportunities to sell for a given amount of content, so I certainly don't rule out monetization as a factor.

For me, it's something of a tradeoff. On the one hand, there aren't many courses that are good enough and that I care enough about to devote university levels of work to. On the other hand, I'm taken some that were so trivial that it would be hard to say I got anything lasting out of them.

I understand that the longer the course is the less people will complete it, but at the same time, if you have to write 3-5 lines of code for each assignment what is the value in there? What does the certificate certify?
No argument here. As I said, I've taken or started to take a few courses that were so cursory (at least with respect to my level of existing knowledge) that they were largely a waste of time. I may not be prepared to spend 10+ hours a week over the course of 2+ months for very many classes--but if the whole class amounts to watching a webinar or reading a chapter in a book, it's probably not very valuable.
Is it worth paying for a course?

No.

Is it worth paying for a course so that you complete it?

Yes. Paying for something hits us hard in the accountability checkbook. The more you pay (and the harder it is to get out of), the more you are committed.

Is it worth paying for a course and telling your friends, significant other, random strangers that you are going to learn X, even if that course is only part of learning X, so that you cannot quit because of pride and accountability?

Big YES.

> Is it worth paying for a course and telling your friends, significant other, random strangers that you are going to learn X, even if that course is only part of learning X, so that you cannot quit because of pride and accountability?

Some people would mentally check-off the task as "done" after making a financial effort.

I learned that lesson when I was 20 and I saved money for weeks (I didn't have a job) to buy Hofstadter's Gödel, Escher, Bach. I told my friends how awesome the book was supposed to be. Then when I finally bought it I never got past the first chapter.

I never got rid of the book so that every time I see it I reminds me of that time. I now know that committing financially, or by telling other people about your goals doesn't mean a lot. Committing yourself mentally to the task is what's important.

I agree 100% with this
> Is it worth paying for a course and telling your friends, significant other, random strangers that you are going to learn X, even if that course is only part of learning X, so that you cannot quit because of pride and accountability?

This would kill the initiative for me. As soon as I get excited and tell someone about what awesome thing I am planning to do the likelihood that I will actually complete said thing goes down significantly. A few years ago I saw a TED talk on the topic ("Keep your goals to yourself" by Derek Sivers) and it made perfect sense. Now until I gain enough momentum in a project to feel confident that there's no turning back I just keep it to myself.

About 5 years ago when I was still in high school I found $500 in a derelict paypal account that had been locked. Having finally gotten a debit card I was able to unlock the paypal account by linking my bank account. I decided to immediately chuck $350 of it at coursera (although it could've been a similar product I can't remember). I wanted to become a successful mobile app developer and self-learning seemed the best way to go.

Long story short the courses I happened to pick were of low quality and never completed. Four years later I still cringe when I hear the word eclipse (even though I'm getting into mobile development now after a few years in webdev land). While I'm sure there are a number of very good courses, and quite a few people capable of completing them and benefiting from them in their lives, I believe I was at the time not only the prime demographic for the product but also in the majority of their users.

TLDR; you can make any learning opportunity beneficial. Is courser and similar products worth the effort to make productive? Maybe, maybe not. There are other alternatives that should always be researched.

OT: As a web developer, you might be more comfortable with iOS development. Checkout Swift and Xcode Playgrounds. Even though Android Studio 2.0 was just released this week, iOS and Xcode is more like Visual Studio (complete ecosystem) than Android will ever be. Good luck!

https://developer.apple.com/library/ios/recipes/Playground_H...

Android Studio 2.0 released https://news.ycombinator.com/item?id=11449029

That really was off topic...
>I wanted to become a successful mobile app developer

>As a web developer, you might be more comfortable with iOS development

How was that off topic?

Or maybe my reading skills are off. I reread it before I posted but still, well deserved ;-)
I agree with some of the comments about effectively being able to play the exam game. That said, some courses I've seen have offered extra credit, and distinction grades, if certain extra work is undertaken which is not simply guessing exam questions multiple times. Such an example is Cryptography by Prof Katz which is part of the Cybersecurity specialisation. You can pass the course by playing the system, but to do well you need to write code to crack crypto systems. Not something everyone is capable of and offers a different perspective on the overall result. I've done several coursera courses and they've landed me some pretty decent jobs. Just depends on how You sell them :-)
It 100% depends what your goal in taking the course is.
If you're talking only about the certificate: No.

Put the skills you learned in your resume or start working on projects which demonstrates those skills. List out the courses you completed under the 'Trainings' section of the resume and Mention MOOCs under your hobbies (If you have such a section a resume).

I was very interested in MOOCs (ai-class was awesome!) but unfortunately most of them seem to have gone the 'pay XX to get this certificate you can to print out!' route to make money.

The fact that you're interested in MOOCs says many positive things about you and shows that you're interested in learning and improving your skill-set, and at this point in time, I would be very surprised if anyone gave two shits about the certificate, because who the hell lies about taking a MOOC?

And in the case of coursera - many of the courses are actually poor quality. Udacity and Edx (especially EdX) seem to take quality much more seriously.

I took and completed Algorithms: Design and Analysis, Part 1 (https://www.coursera.org/course/algo) for free, and it was 100% worth it. It was just the right amount of challenging, intellectually stimulating, and language agnostic so I think it actually made me a better JavaScript programmer in addition.

I made it a point to meet all the deadlines and assignments, which kept me honest and forced me to be timely. It was free when I did it, but I'm not sure if that's changed since looking at other courses I don't seem able to receive a certificate or weekly grades if not paying... Point being it may be worth it to pay just so you feel incentivized to finish the course with a passing grade :)

I've taken several Coursera courses all the way since the beginning. My general impression is that the quality has gone down. Personally, I don't care about the certificates. I think a much better way to demonstrate that you have taken a course is to apply what you have learned, or blog about it, or review it on one of the review sites (I think the biggest ones are Class Central and Course Talk).

Courses I have taken (and reviewed):

https://henrikwarne.com/2015/10/20/coursera-course-review-so...

https://henrikwarne.com/2015/01/01/coursera-course-review-co...

https://henrikwarne.com/2013/02/18/coursera-algorithms-cours...

https://henrikwarne.com/2012/05/08/coursera-algorithms-cours...

https://henrikwarne.com/2011/12/18/introduction-to-databases...

https://www.class-central.com/mooc/4893/coursera-the-global-...

Review sites:

https://www.class-central.com/

https://www.coursetalk.com/

I think this is a great effort. Are you at beginner level or advanced? My nephew is currently learning to code and the book he is using is very much university level content (read: utter garbage), it goes as far as distinguishing "inner loops". Any developer knows that there is nothing special about an inner loop (besides O(N^2)) - it's nothing more than logical progression. I'm looking for something worthwhile for him to learn from.

I've saved these reviews and will be contemplating them for his education - it would help to know how far along you were when approaching them.

Depending on how you structure your loops, things can be optimized on how it hits the cache and reads contiguous memory addresses.
There was a time (late 90s) when GCC would optimize conditionals in an asymmetric way. Somebody will hopefully correct me, but I think the advice was to make conditionals usually be true, for best performance. Since most coders would never hear about this asymmetry (though that's just my impression) it was a bizarre case of favoring one branch of a conditional over another for purely historical (machine code) reasons.
I think now branch prediction on processors can go either way, just as long as it's biased one way or another. If it's not biased one way or another, try to reduce out branches into vectorizable instructions (or pray that GCC takes care of it for you).

Branch prediction isn't a GCC thing, it's a processor thing.

Observe this wonderful Stack Overflow question: http://stackoverflow.com/questions/11227809/why-is-processin...

I did an MSc in computer science, and after that I had worked as a professional programmer for about 20 years when I started taking the courses.
I don't think is worth to pay for a course or specialization beyond the motivation of not losing the money invested because there is no difference at all between the free and paid version, which is more frustrating are the support forums, they are run very cheaply, getting an answer could take weeks and it's just the blind leading the blind, also there is no guarantee that a course offered today will be offered in the future, have that in mind if you start an specialization, if you miss a course it's impossible to get something that you paid for.
The forums and peer-to-peer interactions more broadly are unfortunately one aspect of MOOCs that scale poorly. MOOCs haven't really lived up to their hype (they're valuable but not transformative) because the thing they do really well--broadcast lectures--was largely a solved problem. While the thing that they do poorly--individual or small group interactions--is the part of education that's difficult and expensive.

For certain types of content, which happen to line up well with programming topics, autograding systems are another good innovation. But otherwise peer-grading and multiple choice tests are pretty weak.

I don't think is worth to pay for a course or specialization beyond the motivation of not losing the money invested because there is no difference at all between the free and paid version, which is more frustrating are the support forums, they are run very cheaply, getting an answer could take weeks and it's just the blind leading the blind, also there is no guarantee that a course offered today will be offered in the future, have that in mind if you start an specialization, if you miss a course it's impossible to get something that you paid for.