Friday, November 18, 2005
Hurricane forecasting for the Atlantic basin
Science and Operations Officer
National Weather Service
We will cover factors that affect hurricane development, hurricane climatology, hurricane modeling, and a review of the 2005 hurricane season.
Friday, October 28, 2005
Statistical Analysis of Coral Reef Data
Since 1996, the Florida Keys Coral Reef Monitoring Project has been sampling annually some permanent coral reef sites off the Florida Keys. Three sets of data are obtained: video transects data, species inventory, and absence or presence of disease per species. Using nonparametric Friedman repeated-measure testing, the hypothesis that coral cover remains the same over time was tested sanctuary wide and across regions and habitats. This hypothesis was rejected for both 1999-2004 and 2003-2004 time periods sanctuary-wide, for most of the habitats and Lower Keys region; it was not rejected for Upper Keys and Middle Keys. The investigation of coral cover on station level, using proportionality testing, revealed the stations with no change and those with significant loss or gain over the years 2004 vs. 1999-2003 and 2004 vs. 2003. The analysis of species inventory data, based on a jackknifing estimate of the Shannon-Weiner diversity index, showed a significant loss in species richness. The presence of bleaching, white pox and other diseases has increased over the years from
1996 to 2002, while black band disease has remained practically unchanged. We will discuss statistical data analyses as an efficient tool for assessing spatial and temporal changes in coral reef ecosystems.
Friday, October 14, 2005
Individual-patient data meta-analysis (IPD MA) in the presence of competing risks
H. Lee Moffitt Cancer Center, USF
In conventional IPD-MA of time-to-event data, events of the interest which are yet to be observed are “censored”. Censoring is typically assumed to be “non-informative”, i.e., the probability of observing the subsequent event of interest is not affected by any characteristics of the study patients. However, an alternative view is that if the patient experienced a different event of interest (e.g., death) he/she cannot be at the risk of developing an outcome that we may be interested in. Therefore, these competing
risks (CR) should be taken into account in the analysis of survival data. We performed IPD MA of 9 trials (\(n=1,111\) patients) that compared allogeneic peripheral stem cell transplant (PBSCT) with bone marrow transplant (BMT) in treatment of hematologic malignancies. The event of interest was extensive stage of chronic graft vs. host disease (cGVHD). The analysis was
performed with and without taking two CR (death, relapse) into account. IPD MA of time-to-event data may produce dramatic differences in the results depending on whether CRs were taken into account. The users of such evidence should be aware of this. We provide, for the first time, a method for performing IPD MA in the presence of CRs along with user-friendly program in STATA statistical package.
Friday, September 30, 2005
Avenues for Collaboration in Volcanology and Statistics
Geology Department, USF
Volcanoes are complex physical systems. Worldwide, volcanoes present hazards to many communities and hazard mitigation can only be accomplished through collaboration between scientists with varying expertise. Data on volcanic systems will be presented with a eye toward developing collaborative research projects related to the time scales of volcanic eruptions, patterns of volcanic activity, and geophysical flows that produce hazards at erupting volcanoes.
Friday, September 23, 2005
Multiple Regression Estimation of Sexual Orientation Income Differentials
Multiple regression is one of the most common statistical techniques used by labor economists to analyze and estimate wage differentials between groups of workers. In the last decade a few economists have expanded these techniques to analyze differences in wages between individuals of different sexual orientations. An expansion of this body of work, which accounts for the correlation between marital status and sexual orientation, possible different income effects between bisexuals and homosexuals, and differences in the hours worked across sexual
orientations of women, will be presented.
Friday, September 16, 2005
On Hastings-Metropolis Algorithm
Friday, September 9, 2005
Record Values of Univariate Distributions
Mohammad Ahsanullah (Moe Ahsan)
Record values of univariate distributions are defined. Some basic distributional properties of record values are given. Inferences based on record values for Type I Extreme Values distribution are presented.