I'm trying to find a reliable way to come up with really fantastic interview questions. In particular, technical screening questions you might see during the first stages of a software interview.
I've got lots of ideas on what makes a great interview question great. I've started coming up with questions, but I've quickly found that I need to test them to see if they are any good.
I need to find a reasonably scalable way to get feedback on my prototype questions. Does anyone have thoughts on where to start?
These processes were never proven to result in any above average quality hire.
And they usually have very low correlation to the actual job.
If you do them face to face sorta, they let you have a conversation.
But getting the best quality questions in no way gets you a better hire.
It's a weird thing we do cause we don't have anything better.
Hiring filtering process is just a long list of undescribed biases.
Just have an honest open discussion and see if you jell
I couldn't agree with you more. The sorts of interview questions usually asked in interviews are awful. They do a very poor job of assessing a candidate's fitness for a position. If you expect people to study for your interview, then you're not asking the right questions.
With that said, I also find completely unstructured interviews to be tough. I've learned that just because I like someone, doesn't mean they are a good fit. It also becomes more difficult to compare one candidate to another. Having discussions to determine culture fit are a super important part of an interview, but in my experience, not the only part.
Interview questions give you a structured way to think about a candidate. This is particularly important early in the process. Since weeding out unsuitable candidates early can save a huge amount of time. So I'm trying to come up with interview questions that don't suck. Questions that speak to the experience of the candidate, not knowledge. Questions that kick of interesting conversations. Questions that show the candidate you respect their hard work and experience.
I seem to have gone a little long on this reply :) Anyway, just because most software interview questions are awful, doesn't mean they have to be.
Conversation questions like how would you design X, explain to me what is Y, what would you do if...
Things that are open, provide a discussion and you can follow up further I think provide a better experience both ways...
But it's not about the quality of the question in my mind it's about the conversation.
If both sides know the topics they can read each other.
This test both communication, chemistry and knowledge if done right.
But it's not an apple to apples compare between people, it's about can you work with this guy and will he in high probability get the job needed done.