Wednesday, Aug 6: 2:00 PM - 3:50 PM
4194
Contributed Papers
Music City Center
Room: CC-104E
Main Sponsor
Section on Teaching of Statistics in the Health Sciences
Presentations
A graduate Introductory Biostatistical Methods course is required for all health sciences related Certificate, Master, and PhD programs at the University of Florida. This is a comprehensive course, with all data analysis being conducted using statistical software, and no prerequisites required. Students in this course have drastically diverse background and skill sets which makes teaching this course challenging. We designed this course based on the belief of "learning by doing" and have been offering it in three different combinations regarding the statistical software used and class delivery format: (SAS, in person), (SAS, online), and (SPSS, online). All classes share the same course materials, assignments, and assessment criteria. This study aims at investigating the association between students' learning outcomes and experiences such as quizzes, assignments, projects, perceived difficulty in completing the course, and the software package used, class modality, and individual characteristic factors such as program, mathematical level, prior exposure to statistics, and software experience. Findings from this study will help navigate potential course reform in the future.
Keywords
biostatistics education
public health education
statistics education
analysis of learning outcomes
course design
teaching pedagogy
It was in 1964 that George Box and David Cox published a seminal paper on what now universally labeled as Box-Cox transformation. For a given random variable X > 0, Box and Cox proposed a power transformation of X, which is normally distributed after the transformation. The classical log transformation is a power tranformation in the limit. We query whether the proposed transformation can ever be normally distributed. We demonstrate that it cannot except in the limiting case. We modify the Box-Cox transformation and show that the transformed X now could normally be distributed. The new transformation gives rise to a new class of distributions on the positive real line joining the well-knowm distribitions such as weibull, lognormal, gamma, gumbel, etc.
Keywords
Order Statistics
Optimization
Beta Distribution
Winning Probability
Game Show
Expectation
The purpose of this presentation will be to provide a template for a graduate level program in clinical data science. Drawing upon the historically influential disciplines of statistics and computer science, this program is designed to be responsive to the drastically changing landscape of biomedical research today by offering instruction in the measurement, acquisition, care, treatment, and inferencing of clinical research data. The program contains a series of four core courses (Introduction, Roles & Responsibilities, Design & Implementation, Practicum), five required research courses (Biostatistics, Medical Informatics, Regulatory Science, Clinical Medicine, Clinical Research Ethics), and a list of potential electives. Core features of this 33-hour graduate program include coursework in clinical research ethics, a required, full-time internship in an industry-leading research setting, and the fact that the program is based upon fundamental theories of educational theory and practice.
Keywords
Clinical data science
Data science
Biomedical Research
Regulatory Science
Curriculum
Principal Component Analysis (PCA) is a robust technique for data reduction and clustering analysis, yet its interpretative power in public health data remains underexploited. This study elucidates effective practices for applying PCA in clinical data science, using public health datasets to assist policymakers in interpreting national trends. We analyze data from the CDC Covid Tracker (ED visits and deaths due to Covid-19), the CDC Drug Overdose Surveillance and Epidemiology (DOSE) System (suspected drug overdoses), and the AMA End the Epidemic initiative (buprenorphine, naloxone, opioid prescriptions). These datasets include aggregated measures such as weekly ED visits and monthly percentage changes across all states over at least five years. Through PCA, we reveal latent relationships within the data. Preliminary results indicate that the first principal component may represent a national trend, while the second captures regional variations. By translating mathematical constructs into practical interpretations, we enhance the accessibility of these analyses and support public health policy aimed at reducing geographic health disparities.
Keywords
principal component analysis
PCA
Data reduction
Public health
Covid-19
Clustering analysis
First Author
Zhixin Lun, University of Colorado Anschutz Medical Campus
Presenting Author
Zhixin Lun, University of Colorado Anschutz Medical Campus
A collaboratory is a creative group process designed to solve complex problems that brings the opportunity for new networks to form. This year the Institute for Public Health Practice and Research Policy funded our proposal to establish the Causal Inference Collaboratory at the U of Iowa. This initiative aims to foster collaboration and methodological advancements, positioning our budding group as a resource for researchers to advance causal inference research at our university. Our group has three primary aims: 1. To conduct a review of how causal theory and methods can provide innovative insights into public health research broadly and at our university. 2. Develop a program through workshops and collaborative projects with a Graduate Research Assistant. 3. Create a platform for collaboration and continuous learning through working groups. This component emphasizes collaboration on competitive grants related to causal inference research. This talk will showcase our successes and challenges of working toward these aims, highlighting the outcomes achieved through new collaborations and impactful research. Findings will include characterization of our ongoing research and teaching.
Keywords
causal inference
To obtain a Master's of Public Health (MPH) degree in Biostatistics, students must complete a Capstone project. This requires students to create a committee, determine an appropriate project, conduct a literature review, write a proposal, obtain IRB approval, perform an analysis, write and present a paper. As Biostatistics MPH programs grow, this fantastic experience for students can become a challenge for the faculty who serve on multiple committees each semester. To help ease faculty workload, I created a Capstone course based on replication studies. In this replication course, the students choose from a curated list of publications, with publicly available data, to replicate the results based on their reading of the original paper. The students propose a 1) pure replication, where they replicate the figures and tables in the paper, 2) a measurement and estimation analysis, where they perform robustness checks, and 3) a theory of change analysis, where they perform an analysis to extend the original findings. Students who have taken the course have provided positive feedback regarding their experience. I will provide lessons learned from three semesters of teaching the course.
Keywords
Biostatistics Capstone
Replication research
Capstone Course
Public Health
First Author
Lynette Smith, University of Nebraska Medical Center
Presenting Author
Lynette Smith, University of Nebraska Medical Center