[lsjroberts]

We've been running a very effective initial filter coding interview that simply tests the very basic ability to write loops, refactoring and a little recursion. It's astonishing how many people with CVs of 5+ years professional contract experience in London fail this test. We've had a <15% pass rate on that test alone.

Personally I'm not a fan of taking these interviews, but now being involved on the other side I've seen how essential it to filter out those who don't actually understand how to code.

[danieltillett]

Have you looked into why these people fail? I see all the time claims made like this (i.e. that a large percentage of programmers fail fizz-buzz type questions), but I never see why. Are they morons, frauds, frazzled, confused?

[lsjroberts]

Of the 6 tests I've personally run, 1 passed decently, 2 were probably affected by nerves, 3 just weren't capable. We've had >30 people come in so far and offered to 3 or 4 I believe.

The test is essentially writing a function that can take any number of numbers and add them together followed by a couple of bits around javascript's built in functions and object syntax.

1 of the failures couldn't write a for loop despite having 5 years experience including React.

The test also works as a general personality test as we need people who are fully confident in their abilities due to the difficultly and scale of the product. We are specifically aiming for the top 20% of candidates who come in, so there are subsequent tests after this one.

I thought I'd be able to spot a bad CV, but it turns out either people are lying or they are crediting other people's work to themselves when they are able to hide amongst the team at large organisations.

[coredog64]

That sounds like my experience as well. We have been hiring for senior level positions and ask a fairly trivial string manipulation question (no code, just algorithm) and I have yet to see a candidate answer even close to correct. I suppose one or two candidates could have been nervous, but I expect the rest were truly imposters.

[danieltillett]

Thanks for this. I have not found this such a problem for developers I have interviewed (might be because I am hiring niche backgrounds), but for the sales area frauds are a huge problem. I wish there was the equivalent of fizz-buzz for sales people.

[ionwake]

Did you look at any of their portfolios? Were they taken into account?

[lsjroberts]

It may be a difference between contractors and perms, but it seems not many actually have a github or similar portfolio. We're more interested in their previous experience and if it is relevant (fintech specifically) though.

We probably should check it out for those that have one to be fair. But I'd rather it didn't factor in as a implicit negative for those who are too busy to work on projects outside of the office.

[ionwake]

Thank you for sharing your opinion which is really quite contrary to mine. To me a portfolio signifies where their interests are and what they are clearly capable of achieving without help.