New Paper Explains Success of Factorial Designs in Tobacco Research

Factorial designs are highly efficient and useful in evaluating clinical intervention components, but only when done correctly, according to a new paper in Behavioral Therapy.

The paper, by a slew of UW-CTRI researchers led by corresponding author and UW-CTRI Research Director Dr. Tim Baker, serves as a primer to help researchers determine when a factorial design is appropriate, as well as best practices for using such a design.

Conducting research using a factorial design has many advantages, as it allows one to look at individual treatment components more closely and efficiently. It also helps examine how different treatments interact with each other within a single experiment.

“Researchers may want to incorporate factorial designs into their research because they want to be able to answer important research questions, but they don’t know how to best use it,” UW-CTRI Associate Director of Research Dr. Megan Piper said. “This was our attempt to say, ‘here’s what we’ve learned from our experiences using and analyzing data from factorial designs.’”

Piper explained that UW-CTRI researchers have learned a lot since they first collaborated with Penn State University’s Dr. Linda Collins on factorial designs during the first P-50 study grant.

“We found out that interventions interact more than we thought they would,” Piper said. “We sometimes got interactions that we weren’t expecting, but that is how we were able to learn all the intricacies, pros and cons of using factorial designs.”

Factorial designs originated in engineering research, where researchers could easily alter various components of objects such as widgets and vacuum cleaners. When applying factorial designs to clinical research, however, Piper explains that it is important to be mindful of how many interventions a patient experiences—and whether or not two different interventions make sense when they are combined.

“People bear a burden when you add additional factors, as opposed to when you change various components of something like a vacuum cleaner,” Piper said.

Going forward, Piper said this primer highlights many of the advantages of factorial designs in clinical research while also explaining the considerations that have to happen in order to implement the method in future studies.