Computing graduates deemed unemployable

[epenguin]

We often get questions here on the lines 'I'm really passionate about physics/philosophy/xyz but I/my parents think the jobs are in computing/engineering/medicine/uvw what should I do?

Most often the advice is do what you're really interested in

The other day in 'The Times' (UK) was an article with the above headline.

I didn't know this was a fact, but I can't say I was that astonished.

Headlines often change over a day and in the version I'll try to link to this is toned down to

Geek-speak graduates leave firms at risk of being hacked



"Computer science graduates have the poorest employment rates of university leavers because they struggle to communicate without using geeky language, and learn little about cyber-security during their degrees, an expert claims.

He is backed by senior industry figures, including those from banking and the NHS, who have attacked the poor quality of computer science degrees which they say give graduates little advantage over other university leavers in finding IT work.
They are the least likely to be employed according to figures from the Higher Education Statistics Agency (Hesa), which compared the employment prospects of graduates six months after leaving university in 2013.

Thirteen per cent of computer science graduates were out of work, compared with 8.7 per cent of creative arts and design graduates, 8.6 per cent of maths graduates and 7.4 per cent of those who took degrees in languages."

According to the figures which have limitations but surely tell something it looks like in the UK at lease, the typical pushy-parental beliefs are true for medicine where there is very little unemployment. And even not much for vetinary graduates, which I think is a change from a few years ago; entry must have been restricted or fallen of its own accord. But for the rest it does not seem to help much to do one of the more 'practical' subjects. And language graduates and those in history and philosophy do a little better than 'hard science' graduates, only a few things like meeja studies can really hold you back a bit.

And I am not surprised but let us discuss.

These are figures for the UK. In a lot of other European countries around 10% of graduates not finding a job 6 months after graduation would be an if-only dream.


http://www.thetimes.co.uk/tto/education/article4167779.ece

https://www.hesa.ac.uk/free-statistics

https://www.hesa.ac.uk/images/stories/hesa/Pubs_Intro_Graphics/DLHE_1213/dlhe_1213_table_E.xl



Percentage of students jobless six months after graduating in 2013

Medicine & dentistry 0.2

Education 3.2

Veterinary science 5.2

Law 6.4

Architecture 7.2

Languages 7.4

History and philosophy 8.2

Physical sciences 8.6

Mathematical sciences 8.6

Engineering & technology 8.7

Social studies 8.7

Creative arts & design 8.7

Business & administration 8.9

Mass communications 10.7

Computer science 13

 

[ActionPotential]

This is a very interesting discussion. I am assuming by 'graduates' we are referring to those who have obtained either a Master's or PhD in Computer Science -- is this correct?

 

[epenguin]

British readers to whom the Times article is addressed would I think assume this refers mainly to the 3 year B.Sc. or 3 + 1 year M.Sc. you can obtain detailed breakdowns of employment/degree type if you wade through tables e.g.

https://www.hesa.ac.uk/pis/09/10/emp

https://www.hesa.ac.uk/dox/performanceIndicators/1213_T21R/se1_1213.xlsx

 

[Shaun_W]

These figures are a few years old and have been very hotly disputed. As far as I know, IT was included with computer science, which I'm told brings the average down.

It's also worth noting that those statistics don't include what type of jobs are obtained. Law, for example, fairs very favourably, but in reality only 1 in 5 law graduates ever qualifies as a lawyer.

But given that nowadays most graduate jobs don't care what degree you studied it doesn't surprise me that generally they are all bunched together in the employment stats table, which makes your advice of doing what you're interested in very sound.

 

[JakeBrodskyPE]

For those who have not seen it, I highly recommend reading about the Fizz-Buzz test. There are other blogs such as http://blog.codinghorror.com/why-cant-programmers-program/ with claims of frighteningly common levels of ignorance and incompetence.

I have no reason to dismiss these claims, but it does open the question of what schools ARE teaching. I'm not real sure who is to blame here. Is it that the students are expected to understand something that is not part of most curricula, or is it because teaching the fundamentals and process of programming is somehow beneath the dignity of a computer science department?

There was a time when high schools taught students how to program in BASIC, Pascal, Logo, or some academically oriented language. Learning to program in at least one high level language, learning the process of writing a program, and debugging what you wrote, is a big deal.

But for whatever reason, that is not happening any more. So people flunk the Fizz Buzz test very badly, in terrifyingly large numbers. It's not just because they don't know how to program, but because they don't even understand the nature of integer math.

This problem is so huge, I'm not even sure where to start.

 

[jesse73]

For those who have not seen it, I highly recommend reading about the Fizz-Buzz test. There are other blogs such as http://blog.codinghorror.com/why-cant-programmers-program/ with claims of frighteningly common levels of ignorance and incompetence.

I have no reason to dismiss these claims, but it does open the question of what schools ARE teaching. I'm not real sure who is to blame here. Is it that the students are expected to understand something that is not part of most curricula, or is it because teaching the fundamentals and process of programming is somehow beneath the dignity of a computer science department?

There was a time when high schools taught students how to program in BASIC, Pascal, Logo, or some academically oriented language. Learning to program in at least one high level language, learning the process of writing a program, and debugging what you wrote, is a big deal.

But for whatever reason, that is not happening any more. So people flunk the Fizz Buzz test very badly, in terrifyingly large numbers. It's not just because they don't know how to program, but because they don't even understand the nature of integer math.

This problem is so huge, I'm not even sure where to start.

I really wonder too. Not a compsci grad but I could do "fizzbuzz" in any language I remember the print syntax and supports modular arithmetic. I think the problem might be some people who graduate only learning syntax and are absolutely atrocious at basic math

 

[Shaun_W]

For those who have not seen it, I highly recommend reading about the Fizz-Buzz test. There are other blogs such as http://blog.codinghorror.com/why-cant-programmers-program/ with claims of frighteningly common levels of ignorance and incompetence.

I have no reason to dismiss these claims, but it does open the question of what schools ARE teaching. I'm not real sure who is to blame here. Is it that the students are expected to understand something that is not part of most curricula, or is it because teaching the fundamentals and process of programming is somehow beneath the dignity of a computer science department?

There was a time when high schools taught students how to program in BASIC, Pascal, Logo, or some academically oriented language. Learning to program in at least one high level language, learning the process of writing a program, and debugging what you wrote, is a big deal.

But for whatever reason, that is not happening any more. So people flunk the Fizz Buzz test very badly, in terrifyingly large numbers. It's not just because they don't know how to program, but because they don't even understand the nature of integer math.

This problem is so huge, I'm not even sure where to start.

Programming isn't typically included in the curriculum at schools any more. Most universities seem to want maths and physics to as high a standard as possible for entry, with little mention of computing or programming. I don't study it myself (although I came close to it!) but I've heard that computer science degrees in the UK are essentially applied maths degrees, which could perhaps explain the entry requirements.

 

[jbunniii]

But for whatever reason, that is not happening any more. So people flunk the Fizz Buzz test very badly, in terrifyingly large numbers. It's not just because they don't know how to program, but because they don't even understand the nature of integer math.

I used to assume that applicants knew rudimentary programming and focused on higher level questions. After being burned a couple of times, I no longer make that assumption. Now I always ask the applicant to write a few simple C or C++ functions: print a string in reverse order, count the number of "1" bits in an integer, etc. Also elementary questions like what's the difference between

char *str = "abc";

and

char str[] = "abc";

At least half fail, and this is among experienced software engineering applicants. No way is someone going to touch our source code if they can't even answer these questions, if I have anything to say about it.

 

[Vanadium 50]

We had a programmer search a while back, and while I was not on the committee, I got an earful from people who were. A good three-quarters could not pass the FizzBuzz test, including one person who came in saying that he hoped we wouldn't ask him something so elementary. I didn't understand who someone could be aware that this might be asked and couldn't answer it.

Another one where many people fail - "Write a program that takes a string and turns every p to q, and then undoes it." The second part they turn all the q's to p's - even the ones that started as q's.

 

[homeomorphic]

So people flunk the Fizz Buzz test very badly, in terrifyingly large numbers.

Or encouragingly large numbers, if you are someone looking for a job and wouldn't flunk it.

Anyway, those parents are still right that a computer science degree is better than a PhD in some esoteric junk for which the supply far exceeds the demand. If they can get a PhD, presumably they can learn to pass the fizz buzz test, so evidently they are ahead of the competition and should still be able to land a job, given a solid job-search strategy.

If I had a son or daughter who wanted to go to grad school, I would do my best to convince them it was a bad idea, but if they really wanted to go and could get accepted, I wouldn't stop them. Employment opportunities are far from being the only downside.

 

[mal4mac]

"Computer science graduates have the poorest employment rates of university leavers because they struggle to communicate without using geeky language..."

So why do physical science students do better? I graduated in physics in the UK in the 1980s and usually "got the job" in competition with equally capable CS graduates, for CS jobs. My conclusion - many former physical science graduates are on the interview panels and want members of the same "glorious" club, rather than a bunch of, what they perceive as, upstart computing geeks motivated primarily by money, who are too timid to tackle a really tough subject at degree level. (Note - I don't think this - as I said, I thought the CS candidates I competed with were equally capable...)

 

[Choppy]

There are a lot of variables to consider when looking at data like this. The first questions that come to mind are whether any of the differences between the 8.2 and 8.9 subjects are statisitically significant. Instinctively I would assume that the difference between a 0.2 and a 13 is significant, but I haven't read the study.

But let's extend the benefit of the doubt and assume that in fact, computer science graduates are facing a larger unemployment rate than the other majors.

The issue is to consider the headline "unemployable" next to the datum of a an observed rate of 13%. Wouldn't that mean that 87% of graduates ARE getting jobs? How does that translate to "unemployable?" I don't mean to brush the higher unemployment rate under the carpet, but the attention-grabbing headlne seems a little unwarranted to me. I've actually seen commercials for technical schools where they boast four out of five graduates are working in the field within X months of graduating.

The other issue is for computing science departments. Their graduates are facing higher unemployment rates. So what should they do, if anything? There are mulitple dimensions to assess.
1. Is this an economical issue that the unversities have no control over? It is possible that the sector that typically employes these graduates has had a downturn, more so than the average, and therefore there just aren't that many jobs to go around.

2. Is the curriculum up to date? This needs to be assessed on a school-by-school and program-by-program basis. I would imagine that due to the fast pace of change in this field, it's necessary to constantly revamp the programs - at least more so than in physics. Of course you're going to run into employment issues if you have some obsolete curricula.

3. What is the student population like in the program? Is it possible that CS has a kind of "dumping ground" effect, where students default into it if they can't find another major? Do they have a low entry threshold?

4. Is it easier to graduate with a CS degree through academically dishonest methods (at least on a relative scale) - leading to a larger sub-population withing the graduating cohort that simply does not have the necessary skills that employers expect to come from that degree?

 

[mal4mac]

The issue is to consider the headline "unemployable" next to the datum of a an observed rate of 13%. Wouldn't that mean that 87% of graduates ARE getting jobs? How does that translate to "unemployable?"

Two interesting tables:

Computing:

http://www.theguardian.com/education/table/2013/jun/04/university-guide-computer-sciences-it

Physics:

http://www.theguardian.com/education/table/2013/jun/04/university-guide-physics

The "graduate career after 6 months" figure is looking worse for physics graduates! Even for the top ten, the average is about 20%! So why the discrepancy between the Times and the Guardian?

 

[heatengine516]

Bumping this thread just to say:

All the computer science students in a CS class I'm taking have been complaining that the practical programming techniques/methods they're being taught are outdated and unnecessary (i.e some of them have or had in the past software/programming jobs or internships, and don't believe what they're being taught is useful anymore compared with what they've seen). I don't know specifics, that's just what a lot of them say. A few of them also raised concerns that too much of the material is conceptual and not practical and at the end of the day, nobody's actually learning how to code. The only ones that do are the ones that already have practical work experience or the ones that have enough free time/interest to learn it themselves.

 

[analogdesign]

"Coding" is easy. People can learn it in a two-week bootcamp.

Knowing WHAT to code, and how to design software systems that are robust, scalable, maintainable, and have sufficient fitness to purpose is really hard. Hopefully that's why there are in a degree program.

 

[symbolipoint]

Bumping this thread just to say:

All the computer science students in a CS class I'm taking have been complaining that the practical programming techniques/methods they're being taught are outdated and unnecessary (i.e some of them have or had in the past software/programming jobs or internships, and don't believe what they're being taught is useful anymore compared with what they've seen). I don't know specifics, that's just what a lot of them say. A few of them also raised concerns that too much of the material is conceptual and not practical and at the end of the day, nobody's actually learning how to code. The only ones that do are the ones that already have practical work experience or the ones that have enough free time/interest to learn it themselves.

That is discouraging. Certain other college and university subjects are also not well treated in their instruction; now that has also affected computer programming courses, at least where you have seen. Practical skills are very important. The goal of learning computer programming is usually to be able to apply data processing for practical purposes.

Without knowing the details, what the faults are in your colleague students' instruction is not clear. What is obsolete which they are being taught? What is unnecessary, and why?

 

[mal4mac]

The hard and interesting part of computer science, as analog says, is design. So the important parts of the course are things like "design patterns" and "object orientation". These are probably best learned in a language like Smalltalk, which ('cause the "real world" is crazy!) is not much used in the "real world". If you learn such concepts in a sensible language, they should be (i) easier to learn (ii) easy to transfer to whatever prize turkey of a language they are using "out there" today.

P.S. The Java language was only adopted because the owners of Smalltalk were not prepared to give it away. Java was quickly cobbled together by average progammers to get a job done, and then just because "it was there" everyone adopted it. That's "practice" for you! (And now I hear that Java is waning... turkeys only last till Xmas...)

P.P.S. I avoid programming these days - it's only something you'd get into if you "need a job". Not a pretty subject - like arithmetic at the time of Roman numerals. Give it a few hundred years and it might be a pretty activity. Meanwhile, stick with physics for university courses. See what a "pure", "developed" and "non commercial" subject looks like before getting your hands dirty.

 

[zoki85]

Percentage of students jobless six months after graduating in 2013

Medicine & dentistry 0.2

Education 3.2

Veterinary science 5.2

Law 6.4

Architecture 7.2

Languages 7.4

History and philosophy 8.2

Physical sciences 8.6

Mathematical sciences 8.6

Engineering & technology 8.7

Social studies 8.7

Creative arts & design 8.7

Business & administration 8.9

Mass communications 10.7

Computer science 13

It is worrying engineering & technology unemployment is rather high. I have been without job for more than a year now :(
Do you have data from 10 or 20 years in the past perhaps?

 

[Shaun_W]

It is worrying engineering & technology unemployment is rather high. I have been without job for more than a year now :(
Do you have data from 10 or 20 years in the past perhaps?

Consider the shift in the UK economy away from engineering and technology and onto services. It's not as easy as it once was for, say, an engineering graduate to secure an engineering job than it once was, because there aren't as many of them as there once was. The big growth in jobs seems to be in services, and jobs in this area by and large do not have a preference for any particular degree, therefore it's likely that the unemployment rate amongst different degrees is now a lot more uniform (because your degree subject is less important), rather than skewed towards anyone particular area.

And I'm also sure that many engineering and technology graduates could find work in finance and services if they wanted to, but rather they're holding out for jobs in fields related to their degree.

 

[Shaun_W]

And we must also realize that a degree in a subject does not guarantee a job in that field. Not only that, though, also that the supply of degrees in a field does not correlate to the supply of jobs in that field.

Secondly, the grade requirements to enter a degree in engineering and technology are typically very low. That means that there could be people graduating with such degrees who might not be best suited for employment in that field, too, which won't help unemployment statistics.

 

[jesse73]

So why do physical science students do better? I graduated in physics in the UK in the 1980s and usually "got the job" in competition with equally capable CS graduates, for CS jobs. My conclusion - many former physical science graduates are on the interview panels and want members of the same "glorious" club, rather than a bunch of, what they perceive as, upstart computing geeks motivated primarily by money, who are too timid to tackle a really tough subject at degree level. (Note - I don't think this - as I said, I thought the CS candidates I competed with were equally capable...)

The assumption being that the CS curriculum and scope as well as the relative availability of CS grads hasnt changed since the 1980s.

As others pointed out lumping out IT grads with CS grads is unfair.