Environmental Health SciencesInformation ResourcesAlumniContact Us
School of Public Health, UC BerkeleySchool of Public Health, UC BerkeleySchool of Public Health, UC BerkeleyGroup in Biostatistics

Alan Hubbard

Assistant Professor of Biostatistics


School of Public Health

University of California, Berkeley
101 Haviland Hall, MC 7358
Berkeley,  CA  94720


PHONE: 510-643-6160
FAX: 510-643-5163
OFFICE: 113B Haviland
EMAIL: hubbard@stat.berkeley.edu
photo



Teaching

PH 242C: Longitudinal Data Analysis
This course covers the statistical issues surrounding estimation of effects using data on subjects followed through time.  The course emphasizes a regression model approach and  discusses disease incidence modeling and both continuous outcome data/linear models and longitudinal extentions to nonlinear models (e.g., logistic and Poisson). The primary focus is from the analysis side, but we will also discuss the mathematical intuition behind the procedures.  The statistical/mathematical material include some survival analysis, normal linear models, logistic and Poisson regression and matrix algebra for statistics.  Next taught in Spring, 2009.  Co-taught with Nicholas Jewell.

PH 296-03: Causal Consulting
The course revolves around researchers (students, faculty, etc.) in the School of Public Health that desire advice on the analysis plan (or design) for their studies.  They will present to the class their particular problem, it will be followed by a discussion that defines 1) the so-called full data one would optimally wish to have to estimate the effect of interest (i.e., the notion of counterfactuals), 2) the definition of this effect as a function of the full-data distribution, 3) the observed data, 4) estimators of the parameter of interest from this full data and 5) assumptions necessary for this estimator to yield unbiased estimates of the parameter from the observed data.  It’s a 5-step program that’s guaranteed to change your life, or at least the way you approach analysis.  We  emphasize how to provide proper inference that accounts for the amount of exploration (model selection) one has done with the data.  Next taught in Spring, 2009.  Co-taught with Mark van der Laan.

PH 298-53: Methods in Social Epidemiology
This course is designed to review, evaluate and apply methods currently used in the field of social epidemiology.  The course aims to teach approaches to forming clear research questions, and selecting the best method(s) to answer the questions posed.  Initially we will discuss approaches to defining clear and specific research questions.  We will then discuss recent controversies around the meaning of questions posed in social epidemiology, and the ability of currently used methods to answer questions in social epidemiology.  Finally we will review, evaluate and apply a range of different methods that are or could be used to answer questions in social epidemiology, again emphasizing the types of questions answered by these methods, and their ability to address the challenges to effectively answering questions in social epidemiology.  There will be a mixture of discussion and lecture depending on the topic, with student participation and questions strongly encouraged.  Next taught in Spring, 2009.  Co-taught with Jennifer Ahern.


Research

Clustering Functions

This research has revolved around the apparently simple question:  How many different kinds of patients does many data set contain?  It was motivated by a data set from San Francisco General Hospital (SFGH) on several hundred HIV subjects followed after initialization of HAART.  Subjects were followed irregularly over time and both CD4 counts and viral loads were recorded.   The basic method involves an ad hoc part (smoothing and prediction at grid points, clustering) and a rigorous part (choosing the parameters at each step by cross-validation).  The result is a set of clusters defined by the longitudinal profiles of patients.


Dynamic Models of Infectious Disease
 

More of my work has been focused on infectious diseases and the unique statistical issues that arise when outcome data among subjects is inherently related (correlated).   Part of the work involves using mathematical infectious disease models to investigate the potential bias of ignoring the feedback inherent in infectious diseases.  (
Eisenberg, et al., 2003)

In addition, a recently submitted paper on analyzing the different contributions (person-to-person, person-to-environment-to-person) to the Cryptosporidium outbreak in Milwaukee, we used a novel technique to find the posterior distribution (the estimation distribution) of the relevant parameters in the model.  This involved a combination of profile likelihood methods and a modified MCMC algorithm.  (similar to Hubbard, et al., 2002)

Risk Assessment

With Prof. Mark Nicas on assessing risk from respiratory infections, also incorporating previous work on dose-response.  This work is inspired by characterizing risk of infection (and the efficacy of preventive measures) from bioterrorism or infection of hospital workers in an outbreak. (Nicas and Hubbard, 2002 and Nicas and Hubbard, 2003)

Computational Biology

I have recently completed an initial analysis on Affymetrix data and workers exposed to benzene.  The data (from Prof. Martyn Smith’s lab) consists of 40,000+ gene expressions measured on 12 workers (6 exposed and 6 unexposed matched pairs) in China.    In addition, we are examining a very similar data set on dioxin exposure.

Locally Efficient Estimation

Work on (treatment specific) locally efficient estimation in the presence of potentially informative censoring and confounding.   (
van der Laan, Hubbard and Robins, 2002, Hubbard, et al., 2000, and van der Laan and Hubbard, 1998)



Other Activities



Links

Curiculum vitae

Group in Biostatistics