Shauna Sweet

Principal Scientist

Having access to a mass of evidence is one thing; constructing defensible arguments on the basis of this evidence is quite another.”
– David Schum, The Evidential Foundations of Probabilistic Reasoning

Background

In my work, I seek to rigorously apply a broad range of quantitative methods in an effort to develop novel solutions to complex mission problems. Rather than locate my research interests in a particular domain or even identify a particular “space,” I would instead describe my research interests through the following assertions:

1. All data are indicators which are measured with error. In fact, many of the quantities about which we most want to make inferences cannot be observed directly, and accounting for measurement models is as critical to making accurate inferences and as is the modeling of the underlying structural relationships between core constructs.

2. It is not possible to have “all the data.” There is inevitably a gap between the data we use to develop our methods and the population about which we endeavor make inferences using those methods, particularly in contested environments.  In fact, collection limitations, non-random missingness, and data sparsity are common characteristics of the problems which I find most compelling.

3. Data are not inherently meaningful. We ascribe meaning to those data by virtue of their incorporation into (and selection of a particular) analytic framework. In this way, our selection and application of methods is not value-neutral, and this has important implications for fairness and the ethical deployment of automated reasoning capabilities.

Shauna received her PhD in Measurement, Statistics, and Evaluation from the University of Maryland College Park, under the direction of Drs. Gregory Hancock and Jeff Harring. She also holds an MSc in Survey Methodology from the University of Maryland College Park, and a BA in Sociology from Hamilton College.

Shauna has worked as an applied and operational researcher within and around the Department of Defense for over a decade. Prior to joining Galois, Shauna was Research Director of Machine Learning at Two Six Technologies.

When not at work, you can probably find Shauna on her bike, hiking in the woods, or coloring on walls – all with her daughter, Eloise.