My research is currently focused on developing statistical methods for censored survival data. My research focus is driven by an application to Native American health care systems and health disparities. My first dissertation problem is to examine the effect of unknown system migration in electronic health records data under the proportional hazards assumption. System migration is a term I'm using to refer to the phenomenon whereby patients use multiple healthcare system simultaneously. When we only have data from one healthcare system, we may be missing relevant diagnoses from their visits at another system. My disseration focuses on how we can identify these patients as well as reduce bias resulting from some patients who receive a diagnosis of interest in another healthcare system before they receive that diagnosis in our healthcare system. The second project is a simulation study to assess time scale selection in three ways: 1) the variation of the coefficient of interest, 2) the bias and variation of the covariate of interest under misspecification of the time-variable, and 3) the bias and variation of the covariate of interest when the effect of that covariate does not follow the proportional hazards assumption. The final project is to extend the time-dependent area under the receiver operating characterestic (AUROC) curves to the multivariate case in order to improve a "behind the scenes" aspect of projects 1.
Additional future projects include
You can check out my statistical methodology research here.
You can check out my applied AI/AN research here.
You can check out my applied ADRD research here.
The Grillen lab at UCI is led by Dr. Joshua Grill, Professor of Psychiatry & Human Behavior and Neurobiology & Behavior, Associate Director ADRC, Director UCI MIND; and Dr. Daniel L. Gillen, Professor and Chair of Statistics, Director ADRC Data Management and Statistics Core. The group specializes in research focused on diversity, retention, recruitment, clinical trials, and adverse events as they pertain to Alzheimer's disease and related dementias (ADRD). The group focuses on training early career researchers, post docs, and graduate students in ADRD and applied statistical research.
The SMART group at UCI is led by Dr. Daniel L. Gillen. It comprises of weekly seminar style presentations of ongoing research by students and post docs in the UCI statistics department. The group serves several purposes including, but not limited to, presenting research, listening to research, asking statistical questions, and broadening one's topics of interest. They also give students the opportunity to become familiar with the diversity of research topics within the UCI statistics department.
CAIANH is currently working on understanding dementia in American Indian and Alaskan Native (AI/AN) populations via two main pathways and two main outcomes. On the pre-dementia end, the goal is to understand what sociodemographic factors are associated with onset of dementia and what modifiable risk factors influence the onset of dementia, with a special emphasis on the relationship between diabetes and dementia. After developing dementia, the goal is to understand if (and how) health service utilization changes, what is most effective, and the resulting costs associated with care.
In line with goals of CAIANH and my dissertation research, I am a stakeholder in the midwest and south central region of AIM-AHEAD. My research can be immediately beneficial to Indian Health Services providers in helping to identify patients who may be using other healthcare services and have more inforamion in their health history than is available to the IHS. The goal is to develop AI and ML resources that advance the health of IHS patients.
Through a connection with Dr. Kahlib Fischer, I worked with class groups to think about organizing and running quantitative public policy related research projects. Previous projects I worked with include two different projects on the relationship between lockdowns and COVID-19 cases in both the United States and internationally. Another project focused on how voting laws impact voter fraud. The goal of the course is to introduce students to research in the field of public policy, but the projects are meant to be conducted with a potential for presenting and publishing an article over the course of two semesters.
My research interests are primarily driven by the applications of my work to individuals. Listed below are my main statistical areas of interest with reasons/applications sub-listed below the topics: