Firstly, I certainly agree unnecessary algo questions for a job that doesn't involve algorithmic problem solving on a day-to-day is silly and leads to false negatives.
But the simpler problems give you more than just the basic 'can you code' test - it gives you an insight their into thought process.
For example, seeing how somebody progresses from:-
for(var i = 1; i <= 100; i++) {
if (i % 3 == 0)
console.log('fizz');
else if (i % 5 == 0)
console.log('buzz');
// ah hm both % 3 and %5...
}
To:- for(var i = 1; i <= 100; i++) {
if (i % 3 == 0 && i % 5 == 0)
console.log('fizzbuzz');
else if (i % 3 == 0)
console.log('fizz');
else if (i % 5 == 0)
console.log('buzz');
...
}
And dealing with the fact they have a mistake that needs fixing is always interesting to observe, and gives you insight into how they approach problems.It seems so incredibly trivial and irrelevant at first hand but shows you a lot (assuming they've not seen it before.) I've been really surprised by this.
I suspect the idea of there being a science of this is a bit of a mistake - there's plenty of stuff that involves so many factors that it's really hard to know the really ideal approach. A quick look at the actual state of software engineering studies (30 undergrad students != all programmers) or e.g. macroeconomics should give you pause.
Certainly loads of scope for improvement though, I agree wholeheartedly!
Moreover, leaving people to do these little tests in the team and proposing them to ask questions whenever they want, plus asking them to explain their solutions helps a lot in seeing the social skill I need (communication, openness...)
Of course, it's not 100% safe, but I'd say it's a good heuristic :-)
And from my experience this is the core problem. People debate over questions/fizz-buzz/puzzles/whiteboard/whatever but usually do not invest enough time to determine what would they like to measure with these methods. "
Yes, I agree
Nobody should write FizzBuzz on the board exactly because if you get a person that can't do that your initial filter/selection is wrong. Dead wrong.
Before hiring an architect do you ask him to sketch a house? Before hiring a surgeon do you ask him to do a mini surgery?
Then don't waste my time with fizzbuzz! Shoot a harder question (even a coding question) and this is saving on time and mutual trust.
Roles are often vaguely defined, because it depends on who is available. People lie, exaggerate and game the system in interviews. Neither side wants to reveal their weak points. Figuring out the underlying truth and matching the best available candidate to the best possible job description is not something algorithms are good at.
competency based interviews are standard for junior positions with large companies in the UK. however, this approach often turns interviews into a bullshitting competition.
Here's the more recent, often cited meta-analysis: http://mavweb.mnsu.edu/howard/Schmidt%20and%20Hunter%201998%... (tl;dr structured interview has validity of .51)
Naturally it has an extensive bibliography if you'd like to dive deeper in the topic.
And even more references can be found here: http://www.people.vcu.edu/~mamcdani/Publications/McDaniel%20...
That is also my experience. As a result I generally avoid applying to big companies where I expect a lot of non-technical interviews. If that's how everyone is hired, I don't want to end up in a MENSA club instead of a place where people have strong technical abilities.
I'd extend your wish to "Whole recruitment process should be less of an art and more of a science". The problem is that even the properly conducted interview using established, good practices (like competence based/behavioral interview) can give you max. 50% accuracy in assessing the competencies. What means that additional methods are absolutely necessary.
But even the most accurate methods will lead to wrong decisions when companies don't have precisely and adequately defined what competencies are actually required to perform well in a given position.
And from my experience this is the core problem. People debate over questions/fizz-buzz/puzzles/whiteboard/whatever but usually do not invest enough time to determine what would they like to measure with these methods.