I recommend against panel interviews. As a candidate in my session at the AYE conference said, “I love panel interviews. I get to run the interview. I only have to be “on” for a couple of hours, I get to decide what to say, and I can manipulate the interview so that it serves my purposes.”
Hmm. If you’re currently using panel interviews (more than one person interviewing a candidate at one time), reconsider. I’ve had much better results with each interviewer staking out the one or two areas the interviewer will ask about, and then focusing the interview on those areas (areas such as design ability, problem solving, planning, ability to get along with others, and so on). At the end of the interview, each interviewer has seen a couple of facets of the candidate in depth. The interviewer may have described the same project to each interviewer or several projects — which is information about the depth and breadth of the candidate’s experience.
If you don’t care if the candidates run your interviews, then by all means, continue using panels. But unless you have some sort of exceptional need for a panel that I can’t imagine right now, panel interviews won’t help you discover enough information about the candidate, so you’ll be making decisions blind. And since it’s so hard to fire people, the last thing you want is a blind decision about a candidate.