During a job interview I might ask the candidate some technical questions to get a feel for his knowledge on what we're talking about.
What if his answers are completely wrong? Should I just accept them or should I tell him the correct answers?
From one point of view, the interview is not the best moment to teach something. On the other side, it could be useful for him to understand it's better to say "I don't know" instead of guessing - or even to become aware that it wasn't as he may knew.
Point out the mistake.
Stating that you think the candidate has made a mistake opens up dialogue to gain insight into the candidate's reasoning. Maybe you expressed your question poorly. Maybe he was momentarily confused with another technology. Maybe he misheard you.
Candidates should not be embarrassed by being corrected. Everyone makes mistakes. You may even gain insight into how they handle feedback.
Pretending they've not made a mistake is patronising and unhelpful to both parties.
What if his answers are completely wrong? Should I just accept them or should I tell him the correct answers?
First make sure your question was heard correctly and fully understood. If necessary, clarify things. And make sure you heard and understood the answer as spoken.
Then, if the answer is simply wrong, just reply "Okay" and move on to your next question. Don't indicate if the answer was correct or not. If you sense that the candidate might be on the right track conceptually but is fumbling for the correct words, you might offer encouragement or even a hint. If you get a wrong answer that shows some particular insight, you could follow up to ask about the candidate's thought process.
But interviews aren't quizzes. Hopefully, you are asking questions that require in-depth answers and not just yes/no, right/wrong. You don't provide a score sheet. Your job as interviewer isn't to teach. It isn't to coach the candidate toward better answers.
Your job is to gauge the fit of the candidate for the position being offered. You need to concentrate on your goal, and not spend your time on other tasks.
Additionally, remember that candidates are often debriefed by external recruiters. Thus you might need to change your questions repeatedly to avoid rote answers.
If you should choose to provide the correct answers for some reason, make sure you do the same for all candidates. You shouldn't pick and choose or discriminate.
And as @PoloHoleSet writes, if you must remark on answer errors at least don't interrupt the interview with a running score of Right and Wrong answers. That could easily fluster a candidate and ruin their concentration.
In technical interviews, pointing out errors is not uncommon. In fact, the interviewer will often say there is a problem and invite the candidate to look again and see if he can find the the mistake. If they don't see it, then guiding them to find the issue is often done.
If the candidate doesn't take criticism well or refuses to work though the issue, then that has given the interviewer valuable information on how the interviewee is like to work with people if they were hired.
I've been in interviews where I got things wrong and the interviewer did point out that I'd made a mistake. I've liked those better than the ones that had just silence. Remember, for job interviews, the candidate is also interviewing the company. Job interviews are not school test time. Companies that run interviews like that are doing both sides a disservice.
Note, I've been hired before when the tables were turned, and I spotted when the interviewer made mistakes on the technical questions (unexpected infinite loop on a tech question).
Tell him.
In discussing the answer, you can lead the conversation into a discussion on why the candidate chose that answer. It could well be completely wrong, but the decision process that went into it could lead into some insights that put the candidate into a different light.
I once sat in on an interview where the interviewer asked the candidate
What's the difference between SQL and Sequel?
Even though there isn't a right answer, the way the candidate answered was nice and displayed a lot of tact.
Basically, you're having a discussion here. It's not just a question of questions/answers.
What if, from the perspective of the interviewee, the answers are not wrong? I once was on an interview which failed to generate an offer or further interest, and I have a feeling it was because of my response to the question "How do you deal with fire fighting?" My response basically was that in my experience fire fighting was the exception not the norm.
That interviewer went on to explain how wrong I was because IT is all about fighting fires. From that interviewers perspective, I was clueless and inexperienced if I didn't know my job was primarily to fight fires. But what the interviewer didn't know was that in my experience, as of that time in the mid 90's, I had built IT services and put processes around them that was far beyond what most organizations were doing at the time.
The interview ended there, and I always have regretted not being experienced enough (at interviewing) to take charge and turn that interview around. Although I probably ended up better off not working there.
Because of this experience, I have always made it part of my own candidate interviewing process to not dismiss "wrong" answers but instead explore it and understand their perspective. Because it is entirely possibly that the answer you are so sure is right is actually not so right.
Assuming you still see some potential merit in the candidate (even if not for the present need or at the current point in their career), you could try to ask additional questions which would lead the candidate to realize the error themselves.
This can also be a part of a general pattern of requesting not only the answer, but asking the candidate for an explanation of the merits and decision logic, and even potentially disadvantages of their offered answer. Ideally you'd do this even when you agree with their basic answer, as how they arrived at it may be as interesting as the fact itself.
But beware of taking this too far:
The candidate may not be in the habit of thinking about problems in the manner required for the position
Once feeling off-base, they many not be doing their best thinking, so even if they'd realize the answer in another setting, they might not in the interview - in which case persistence will only make things worse
There's also the risk that the answer you seek is too obscure and only relevant in context that hasn't been fully explained. This is more a concern when the candidate's answer is not wrong in general, but rather not the answer to the problem you were seeking. Be particularly wary of this in the case of both artificial questions, and also ones that arise from very real problems of project in which you are deeply immersed; asking good technical interview questions is harder than most realize.
And of course if you've already decided the candidate is not a fit and not someone who's future career you want to make a brief contribution to shaping, you could just efficiently and politely move on to wrappings things up.
It depends on the portion of the technical interview. If you're asking questions they should simply know without fault, I might just move on, and if it's a particularly important failure I might revisit it later from a different angle to make sure it wasn't a simple mistake.
During the technical portion meant to discover how they work through a problem, how they collaborate, what questions they ask, etc I'll provide them with feedback of varying kinds.
I might simply correct them outright, "That's wrong," with no further information to see how the respond, giving them a chance to either fix their error or convince me otherwise.
I might ask, "Are you sure?" and see if they can revisit the model they've built in their head, the information they've been given, and find the error themselves.
It's worthwhile sometimes correcting them as well to see how quickly they pick up on the error when given the correct information.
I prefer to treat them as I would treat a co-worker, primarily gauging how they interact with people particularly when there are conflicts between what they believe is correct and what someone else says is correct.
Are they simply a "yes man", do they respond well to criticism, are they able to stop in the middle of an explanation, discuss a tangent, and get right back on topic, do they resolve conflicts with care and compassion, do they have a deep enough understanding to be able to discuss foundational principles and support higher level concepts, etc.
This provides significantly more information than a solely technical interview.
What if his answers are completely wrong?
If you give the candidate feedback is completely up to you, if you feel that giving it would help him/her in some way go ahead, or if you feel that giving feedback will put him better on track with the interview process (as some candidates tend to get nervous on the first questions).
Now, if some candidate you see that is answering all question in an incomplete/incorrect way then it could be a bit rude to make that person realize he is wrong in everything.
In this case what you can do is end the interview earlier, as you now know that candidate is probably not fit for the role. You can give the candidate some feedback on the first questions, but if he continues erring then I would suggest doing what stated before.
When a candidate gives a completely wrong answer, there are three main possibilities:
Stating that an answer is wrong, and even moreso explaining what you expected to hear is in my opinion a mistake. There is nothing to gain from your point of view, but you may very well have disadvantages from doing so. If nothing else, you lose time and may discourage an otherwise acceptable candidate. Or, you will find the next candidate (who, in reality, is none better, maybe worse) being prepared to answer this exact question the way you expect.
A better approach would be to ask the candidate to explain his answer. You can learn a lot from the explanation even if the answer is still completely wrong.
You learn not just the amount of facts and trivia that the candidate has learned by heart, but you learn about his personality and mental ability, in particular the ability to reason, explain, and to adapt to situations.
Note that the ability to store facts and trivia is not the only, and not even the primary thing that makes a good candidate.
The candidate might explain well, or he might get nervous, he might feel offended for having his answer challenged. He might completely lose it, or he might tell you "Alright, you got me there!".
The candidate might give a wrong, but otherwise entirely reasonable answer. Or, he could give an explanation which justifies an answer that is, from a different point of view, perfectly correct -- only just not the one you had in mind. He might discover, and admit, that he has made a mistake half the way down.
This can tell you a lot about whether or not this person will be a good fit by personality, and ability/skill (not dry knowledge). Which is what matters more.
Of course, it doesn't hurt to have the candidate explain one or two correct answers as well. If for no other reason, then to not give out the clue that an answer was wrong.
If the interviewee is confidently giving you completely wrong answers to most of your questions, I would assume you won't want to hire them anyway - so why waste your time trying to teach them something? Either continue the interview just for its entertainment value, or wrap it up as quickly as you can. You don't have any obligations to be "kind" to the interviewee - most of the applicants are going to be rejected for one reason or another anyway!
On the other hand, if you get a completely wrong answer to just one question, the interviewee might have misunderstood the question. Asking "can you explain your answer in a bit more detail, please" is a good way to continue - and if it becomes clear he/she is answering the "wrong question", you can then explain the question better and let him/her try again.
As people have already said, an interview is a 2 way process. As a consequence, we can ask the question:
If you were being interviewed - would you highlight errors said by the interviewer?
To which the answer is not a yes / no answer, but 'does it impact their sutability for the role or the question asked' or 'does raising it have value'.
For example someone states that 5+5 = 9, but then goes off and demonstrates far more complex maths; then you can be pretty sure that they're just trying to get to the answer, and made a stupid slip. Pointing that there was an error and letting them find it (or even pointing out to them directly) would seem pretty reasonable.
If on the other hand, something contentious is said then you get something to discuss, this is a good thing as you can find out what it's like talking to the person, how they deal with confrontation etc etc.
However if something is said that will make them clearly not suitable for the role (or demonstrate that the interviewer has no clue what they're on about) then you either end the interview, or continue to the next question because expanding the point has no value.
An interview is a conversation not a quiz. In general, there are no right or wrong answers. Where there are right/wrong answers we must be open to the possibility that the candidate is right and we are wrong.
We should give the candidate a reasonable opportunity to prove us wrong. We should listen with an open mind. If the candidate makes a strong case - even if we ultimately do not change our minds - we should be impressed.
This is an interview. You want to know something about the candidate's skills as a worker, not as a dictionary. A straightforward correct answer tells you that the candidate is well-prepared, not that he is good.
This is your chance to find out about the candidate's actual problem-solving skills. Provide him with more information and let him work out what is wrong. Of course there is no point in just letting him flounder for the rest of the interview: the real skill is getting fast to the point where he can regain his footing and without turning this into a drama where you are more testing the ability to focus under attack than under work conditions.
