Too long didn't read summary: Ask open questions, provide people with code, talk to them how they would find answers or improve it. And spend a little bit of your time learning some IT if you expect to hire anyone decent.
The problem is that there are many limiting factors in questioning surveys and what you are testing them on. And determining a person's aptitude for technology is a very subjective process. Do you test their resourcefulness by asking them what sites they visit and learn from? Do you test their actual programming ability by having them do a live coding that could be too specific case? Do you ask them about terminology which just shows they may know some buzzwords?
Google for a while gave up on asking algorithm questions supposedly for a certain subset of jobs they had.
I think as IT is getting more specialized and there is a difference between people knowing certain things it is important to have some form of domain knowledge which can be in the form of terminology quizzed, but at what level is up for debate. But maybe a more open question format would be better.
I think always presenting the interviewer with the code and having them look it over then provide input is maybe a better fit. It is like math, everyone wants to just use a calculator but in school you use your head at first for a while, but eventually everyone in a work environment uses a calculator anyway.
The computer questions should be the same, they should have access to a computer. They should then be quized on how to reason out this x code and how can I change it to be more like y. They need to be problem-solving questions, that use some critical thinking but also involves some reasonableness to them. Some may ask terminology, but it shouldn't be the main focus.
It also does not help that employers don't know what to ask for their tech jobs as well in many cases. If someone wants to hire a web developer they may just copy and paste and re-write the job requirements of a random posting they find online. This is what I imagine happens 90% of the time. The reason why job posts are such crap is because when someone posts a new requirement they probably all just update their templates that they had before-hand. It's technological ignorance at best!
What is getting worse is also specialization in different sectors of computing is just confusing companies more-- Devops, machine learning, security programming, etc etc. And as computer literacy is improving in population yet code is getting automated more and more there is becoming an increasingly difficult question of what is necessary to learn for aspiring programmers who will be joining the workforce.
A few good devop team can easily be more productive then 3 times the size of it in developers if the developers know nothing about devops.
A few good machine learners can solve real problems with data mining that may save a big company millions that a group of programmer interns wouldn't know how to do.
A few good pentesters can find holes you completely missed and need to patch providing your company from losing tons of credit cards through an open point of sales or losing important company data.
One funny way to see experience in IT security is to simply ask are we getting more safe, or less safe security wise. The answer less safe shows that they have been working in the IT security field, however if you ask researchers at universities they may not know about the level of problems in the security industry being so prevalent but they may still be good fits for specific research.
It is hard to manage which are necessary, and it is also hard to address skill level. There are many groups of interns now who could easily possess such skills that may be of way more use than people know as well.
Many companies rather hire a student out of college than a phd student because they may have some interest in some of these things and because he may have some sample projects or notice in communities even though he may not have specialized skills.
Hi Eldelshell, your blog shows, that we're lucky to have you in our new team as you also regard things from an abstract way as much as from the material scope. Seeing forward to meet you, "A PM-ing Developer". ;-)
2 comments
[ 3.0 ms ] story [ 15.1 ms ] threadThe problem is that there are many limiting factors in questioning surveys and what you are testing them on. And determining a person's aptitude for technology is a very subjective process. Do you test their resourcefulness by asking them what sites they visit and learn from? Do you test their actual programming ability by having them do a live coding that could be too specific case? Do you ask them about terminology which just shows they may know some buzzwords?
Google for a while gave up on asking algorithm questions supposedly for a certain subset of jobs they had.
I think as IT is getting more specialized and there is a difference between people knowing certain things it is important to have some form of domain knowledge which can be in the form of terminology quizzed, but at what level is up for debate. But maybe a more open question format would be better.
I think always presenting the interviewer with the code and having them look it over then provide input is maybe a better fit. It is like math, everyone wants to just use a calculator but in school you use your head at first for a while, but eventually everyone in a work environment uses a calculator anyway.
The computer questions should be the same, they should have access to a computer. They should then be quized on how to reason out this x code and how can I change it to be more like y. They need to be problem-solving questions, that use some critical thinking but also involves some reasonableness to them. Some may ask terminology, but it shouldn't be the main focus.
It also does not help that employers don't know what to ask for their tech jobs as well in many cases. If someone wants to hire a web developer they may just copy and paste and re-write the job requirements of a random posting they find online. This is what I imagine happens 90% of the time. The reason why job posts are such crap is because when someone posts a new requirement they probably all just update their templates that they had before-hand. It's technological ignorance at best!
What is getting worse is also specialization in different sectors of computing is just confusing companies more-- Devops, machine learning, security programming, etc etc. And as computer literacy is improving in population yet code is getting automated more and more there is becoming an increasingly difficult question of what is necessary to learn for aspiring programmers who will be joining the workforce.
A few good devop team can easily be more productive then 3 times the size of it in developers if the developers know nothing about devops.
A few good machine learners can solve real problems with data mining that may save a big company millions that a group of programmer interns wouldn't know how to do.
A few good pentesters can find holes you completely missed and need to patch providing your company from losing tons of credit cards through an open point of sales or losing important company data.
One funny way to see experience in IT security is to simply ask are we getting more safe, or less safe security wise. The answer less safe shows that they have been working in the IT security field, however if you ask researchers at universities they may not know about the level of problems in the security industry being so prevalent but they may still be good fits for specific research.
It is hard to manage which are necessary, and it is also hard to address skill level. There are many groups of interns now who could easily possess such skills that may be of way more use than people know as well.
Many companies rather hire a student out of college than a phd student because they may have some interest in some of these things and because he may have some sample projects or notice in communities even though he may not have specialized skills.
O...