Transcript

How to get started with design of experiments (DOE)

ANNE MILLEY: What would you recommend for someone that is getting started in experimental design, and what are the biggest obstacles in getting an organization to adopt experimental design methodology?

JULIA O'NEILL: Yeah, it remains a challenge. I think one of the things that holds people back is they feel like they have to get it perfect before they can start. I'd say, dive in and try it. You can start fairly small and simple. My favorite teaching example was actually on a very simple process, but where without design of experiments, we would not have been able to solve the problem. Actually, using design of experiments not only solved the problem, but it led us to developing a mechanistic explanation for what we were seeing, which was a big deal for manufacturing.

The reason only DOE would work for that is that there was a very important interaction. Without using a design of experiments factorial type of approach, I'm positive that interaction would have been missed, and the full answer to the problem would not have been found. I have seen that many times through my career, and I just heard a great story about how people become adopters of design of experiments.

Somebody I've been working with recently said that (he’s a scientist) he was very skeptical about DOE, but he had a boss who forced him to go to a few days of training. He suffered through it, still very skeptical, but he decided to put it to the test.

So he went back to work. They had a problem they had been able to solve after about six months of effort. He said, all right, we're going to pretend we know only what we knew six months ago. We're going to design a DOE and see what it does. So the gauntlet was down. They ran it, and they got the answer in one experiment.

ANNE MILLEY: They didn't have to wait six months like they did.

JULIA O'NEILL: From that point on—I really enjoy working with him, because he's very enthusiastic about statistical methods generally and DOE in particular. I'd say try it, and I think a lot of the reservations people have about it evaporate once they see that it's not actually that complicated.

ANNE MILLEY: It's a shortcut for gaining real insight.

JULIA O'NEILL: I think because you have to put in the upfront planning, people perceive it as being more work and more effort. But in the end, it saves so much time.

ANNE MILLEY: Yeah, a little bit of thinking up front can go a long way. That's a good story. Thank you for sharing that.