Jingyi Duan and Kimberly Hochstedler – both Cornell doctoral students in the field of statistics and data science – each received the International Biometric Society Eastern North American Region’s (ENAR) Distinguished Student Paper Award at ENAR’s spring meeting in March.
Here’s more information on Duan’s and Hochstedler’s award-winning papers:
Jingyi Duan (advised by Yang Ning)
Paper: Two-stage Hypothesis Tests for Variable Interactions with FDR Control
In this paper, Duan and collaborators consider the hypothesis testing problem for interaction parameters, such as gene–gene and gene-environment interactions in genome-wide association studies. To reduce the computational cost and maintain high statistical power, the group propose a two-stage FDR control algorithm for testing interaction parameters with high dimensional data. Duan and coauthors prove that the two-stage procedure provides a valid control of FDR. In addition, the asymptotic power of the FDR control procedure is established.
“In terms of application, we apply our method to a bladder cancer data from dbGaP where the scientific goal is to identify genetic susceptibility loci for bladder cancer,” she said.
Kimberly Hochstedler (advised by Martin Wells)
Paper: Statistical inference for association studies in the presence of binary outcome misclassification
The paper describes a new statistical method that can be used to understand the association between a predictor of interest and an outcome of interest, even if that outcome is “misclassified” – or measured incorrectly – for some of the data. In the paper, Hochstedler considers outcome variables that can take on one of two categories, like a diagnosis, or a yes-or-no question. In this case, “misclassified” means that an observation was in the wrong category.
“In the paper, I applied this method to heart attack diagnosis data,” Hochstedler said. “The goal was to better understand risk factors for heart attacks while dealing with misclassification of the outcome, heart attack diagnosis.”