Crafting Innovative Approaches for Analyzing Data – Dr. Zachary K. Collier

This post was originally published on this site.

Words of wisdom/advice for new faculty: ‘Every obstacle, no matter how daunting, can be overcome. Stay persistent, stay grounded and always leave the door open for others to follow.”Career mentors: Dr. Thomasina Lott Adams, University of Florida; Dr. Bianca Montrosse Moorhead, University of Connecticut; Dr. Walter Lana Leite, University of Florida
Education: B.S., Special Education, Winthrop University; M.S. Measurement and Statistics, University of Florida; Ph.D. Measurement and Statistics, University of Florida
Age: 34
Tenured: No
Title: Assistant Professor, Research Methods, Measurement and Evaluation Program, University of Connecticut, Neag School of Education

Dr. Zachary K. Collier, an assistant professor at the University of Connecticut, focuses on previously unexplored ways to find depth and details. He oversees the Methods for Unstructured and Difficult to Use Data (MUDD) lab to research innovative analytical methods for the social and behavioral sciences, which can contain difficult to use (muddy) data.

Collier is an expert in the field of causal data mining and missing data analysis, which can be used to discover valuable relationships in data that can help with decision making. Key areas of interest for the MUDD lab are developing and applying machine learning and statistics to data in education and social policy and bridging scientific disciplines to combine analytical tools to tackle problems that cross traditional boundaries.

“Dr. Collier’s work is innovative with real-world applications,” says Dr. Bianca Montrosse-Moorhead, a professor of psychology at the University of Connecticut. “His interest is not solely on advancing quantitative methods. His work is also purpose-driven—he uses and tests these methods to study and improve educational and health outcomes. The number of articles he has published and the amount of grant funding he has secured as an early career researcher is exceptional.”

Collier’s graduate studies provided his entry into the field of measurement and statistics. When exploring a topic for his master’s thesis, a senior scholar suggested causal data mining as an emerging field. Collier soon realized he was drawn to data and analysis, which ultimately led to his master’s thesis, his first publication and his academic career. While his original career goal as an undergraduate was to become a special education teacher, he has shifted to a research-focused career, with much of his current work connected to special education and public health.

“I have a Spencer Foundation grant right now,” says Collier on the project titled, Advancing Anti-Racist Special Education Practices and Policies via QuantCrit Methods and Analyses. “It’s looking at overall representation and framing special education as a modern way of resegregation. My component of it is the QuantCrit methodology (rectifying quantitative methods through critical race theory). Coming up with methods to better be able to identify biases and the racist practices that are happening in schools.”

Collier says he blessed to be able to partner with other scholars as he explores new ways to analyze data. Currently he and Dr. Valerie Earnshaw lead a five-year, NIH-funded project aimed at enhancing physicians’ knowledge and skillset in specialty medicine through targeted interventions.

“It can particularly identify stigma in those [future] medical doctors who have biases towards patients, such as men who have sex with men, patients who are female sex workers and also patients who have a history of substance abuse,” explains Collier.

In the MUDD lab, they also focus on creating ways to analyze and handle missing data. This includes things like omitted variable bias. He also explores new ways to analyze and interpret data. The lab was formed when Collier was on the faculty at University of Delaware, and he has kept the same team since moving to University of Connecticut in 2023.

“Much of my new methodology stems from addressing real-world data challenges,” Collier says.

Collier conveys that message to the doctoral students that he teaches. For 2025, he has been fortunate to create two new courses. One will focus on Monte Carlo Simulation (a computational method that uses algorithms for random sampling), which is important for testing a method being created. The other class, which he will co-teach, will be Statistical Learning and Data Science.

“All my classes require a final project,” he says. “A lot of students come in nervous, but by the end of the course not only have they successfully completed the class, but they also have been able to apply a method that they learned in the class to data that they’re interested in, analyze it and have a publication or a conference article.”