I have interviewed between 50 and 100 people over the years, all of them for technical positions, and have been interviewed about 10 times. Its not easy to evaluate a person’s aptitude for a job based on a 30 minute job interview, but I’ve finally worked out an approach and a set of questions that seem to work pretty well. Not surprisingly, my approach seems to be similar to the one that the other interviewers used to evaluate me.

Here are the questions I typically ask, and the answers I’m looking for:

In a nutshell, I’ll explain that I’m a biologist by training, with a Ph.D. and training in plant molecular biology. I’ll describe how many years I’ve been at the company, which divisions and roles I’ve worked at, and what my role is in my current job. Finally, I’ll explain how I expect to work with the person we hire for this job.

This is a useful “icebreaker” question, as it helps the candidate to understand roles and relationships. The most important thing for you, the candidate, to do at this point is to listen carefully and respectfully. How you listen is as important as anything you have to say.

This is a fishing expedition. You should give a to-the-point answer, something along the lines of “this job will give me an opportunity to practice and develop my skills at XYZ, and to contribute to the bottom line of a great company like yours.” Try to give an open-ended answer, and not box yourself—you really won’t know where you want to be within the company until after you’ve already been there awhile. A short, 2-sentence autobiography wouldn’t hurt at this point, but don’t drone on too long.

Explaining technical stuff to non-tech focused people is a huge part of any tech job, and many people do it poorly. I’m not so concerned about your ability to communicate expert-to-expert; I want you to communicate to ME. I will typically find something interesting on the resume (an undergraduate research project is great for this), and ask for a more complete explanation. Start by framing the context of your work, briefly describe the technology, and summarize what you learned or accomplished. You get extra points for correctly gauging my level of understanding by reading my body language, for drawing a simple diagram on a white board, and for coming prepared to talk about this subject be bringing simple props.

One thing NOT to do answering this question, or any other, is to break confidentiality. If you give away your old company’s secrets to me, you’ll give away our secrets after you’ve worked here.

There are difficult people to get along with everywhere. I’ll assume you are not one of them. What I want to know, here, is how you have worked past problems with someone else. This is a good psychology question, as both the choice of problem as well as your problem solving approach offer good insight into how you work with people.

One thing not to do answering this question is to name names. I don’t want to hear specific trash about specific people I may know; I just want to understand how you’ll get along.

My company makes lots of big, complicated instrument systems, which we in product development often work with in the breadboard stage (i.e. components bolted down to a breadboard, not the final product). A little bit of mechanical aptitude and machine-sense is critical. Whatever you answer, though, make sure you actually understand the technology involved. If you only pushed the “on” button, pick something simpler—whatever you bring up will likely lead to follow-up questions designed to probe your level of understanding. Even better, if you are working with complicated systems, take a little extra effort to understand them beyond the surface level—they will be great opportunities to show off your technical wizardry.

Problems are job security, if you know how to solve them. Getting frustrated and calling tech support can only take you so far. I want to hear how you defined the problem, scoped it, hypothesized a solution, tried out the solution, and moved on.

A variation on this question is to give a little quiz: Here is a diagram of a complex apparatus. When you come in one morning, your expected result isn’t waiting. How do you figure out what happened? To answer this, don’t make it more complicated than it needs to be, and don’t feel as though you are supposed to recognize the apparatus. Ask questions: what gauges are present, and are they giving normal readings? Are all the fittings in place? The problem will usually be something obvious, and accessible through simple questions.

I’m a biologist, at a company that makes tools for biologists, so even if I’m interviewing for a software engineer, I ask this one. I want to know that you have passion and curiosity, as well as to find out how much understanding of biology you have.

There are many great answers here—system interactions, subtle levels of medical discovery, emerging technologies. There are also bad answers: the human genome project was cool, and my undergraduate research project is the only interesting thing I can think of. Try to convey the big picture. And this question may be the one in the whole bunch that the interviewer asks genuinely looking for a novel perspective.

So, I came across this one in a newspaper article, and it captured my imagination. No, I don’t know or care what the true answer is; this is about estimation and thinking on your feet. In my job, I need to come up with numerical estimates, in a hurry, all the time. A good answer involves thinking out loud, first establishing assumptions (fuselage only, lets assume no seats or bulkheads), guessing sizes and volumes (lets see, one ball is about 1 inch in diameter… Plane is maybe, 120 feet long? 20 feet in diameter?), then doing mental math to estimate an order-of-magnitude answer (10x10x10 gives about 1000 balls in a cubic foot… Pi R-squared is 3.14 times 10 squared…).

With this or any other outlandish question, the important thing is for you to TRY. The worst answer is a wild guess (“I don’t know, a million?”). If you ask for clarifications, state your assumptions, and try one step at a time, you’re doing great.

Interviews are a 2-way street. A job candidate is evaluating the job at least as much as the company is evaluating the candidate. I’ve had more than one interview where I’ve been invited back, but have said “No thanks.” You should always ask some questions, and you will always have a chance. Some things to ask about are what the management is like, how formalized product development processes are, what relationships between business units are like, what the job will involve, and what life at the company is like.

Finally, there are a few pieces of interview advice that are so obvious you shouldn’t even need to be told:

Be on time.
Dress well, but not too well. Nobody wears suits anymore.
Assume there will be a drug test if they make you an offer.
Show interest and respect.
Don’t apologize for yourself.