Liam Mueller

Research

Biologists have unprecedented access to vast datasets spanning cellular function to ecosystem composition. However, we don’t always have the correct tools to unlock the story these data tell. I focus on working with other scientists in their fields to:

1. Continue designing methods for the interpretation and visualization of untargeted metabolomics and metagenomics.
2. Help develop standardized methods for the comparison of paired matrices data from RDA, PCNM, co-inertia, and others.
3. Find unique and challenging opportunities to apply new statistical techniques to complex biological questions.

As a statistician I understand the limitations of using traditional statistical analyses for education research, and as an educator I also understand the impossibility of a flawless experimental design in a classroom setting. However, I see this issue as a problem for statisticians to solve, rather than a reason to dismiss quantitative education research.

I believe that one solution to analyzing the complexity of quantitative classroom data comes in the form of Bayesian analysis. Hierarchical Bayesian analysis (HBA) can not only improve the precision of estimates, but would remove common complaints quantitative education research receives including unrealistic power requirements, lack of proper controls and debates about appropriate degrees of freedom in complex factorial experiments. My goals are to:

1. Design an R package using HBA with the classroom in mind.
2. Develop teaching workshops and online resources to lower the barrier to entry for HBA.
3. Work with researchers who are excited to try HBA for their projects.

In addition to the specific areas above, I am always excited to work with and talk to people about their statistical questions from any discipline.