UT Wordmark Primary UT Wordmark Formal Shield Texas UT News Camera Chevron Close Search Copy Link Download File Hamburger Menu Time Stamp Open in browser Load More Pull quote Cloudy and windy Cloudy Partly Cloudy Rain and snow Rain Showers Snow Sunny Thunderstorms Wind and Rain Windy Facebook Instagram LinkedIn Twitter email alert map calendar bullhorn

UT News

Life is filled with patterned uncertainties

Life is filled with patterned uncertainties. Supporting evidence is all around us.

Two color orange horizontal divider

Life is filled with patterned uncertainties. Supporting evidence is all around us.

Patterned uncertainties draw the attention of astronomers, geologists, chemists and biologists as well as demographers, health care providers, economists, policy wonks, sociologists and psychologists – to say nothing about university admission officers, political campaign strategists and athletic directors. Patterned uncertainties, buttressed by data, become estimated probabilities. These probabilities are used in a wide range of endeavors to build models of events and processes infusing and defining life.

Several years back, Dean Mary Ann Rankin, of the College of Natural Sciences, recognizing the broad-based importance of patterned uncertainties and estimated probabilities, proposed a new, university-side academic unit designed to support and advance these statistics-employing, model-building endeavors. As a result, the Division of Statistics and Scientific Computation came into being this past year.

With the support of seven deans and the provost, as well as strong encouragement from the Graduate Student Assembly, an ambitious endeavor has been started. New undergraduate and graduate courses have been developed. More are in the works. Consulting for faculty and students has been strengthened. An annual summer statistics institute, offering some 500 seats in introductory and advanced courses, has been established. Faculty committees are convening to organize new graduate and undergraduate concentrations. These may eventually turn into majors. A jointly funded proposal, involving four deans and the provost, is in the works to broaden the faculty expertise and identify a well-focused, university-wide research agenda. It has been a very busy first year.

I recently sent an email to faculty colleagues across campus asking for examples of how estimated probabilities and the construction of predictive models were important to their work. Responses were as varied as they were plentiful. Turns out, statistics and scientific computation are fields few can afford to neglect.

  • Biology, the Brain and Biomedical Engineering: The 1990s were declared the decade of the brain. In 2003 the Human Genome Project was completed. Combined with massively expanded computational power, we can now explore, in detail not previously possible, how we see, hear, taste and feel, how we learn and remember, and how we inherit traits that make us unique. As a result, the inter-related fields of computational biology, bioinformatics and biomedical engineering have grown and many would say have moved center stage.

In 2004 the Center for Learning and Memory was established under the leadership of Professor Dan Johnston. In the past year, three of our faculty, Rick Aldrich (Neurobiology), Bill Geisler (Center for Perceptual Systems) and David Hillis (Integrative Biology) were elected members of the National Academy of Sciences (NAS). Professor Nicholas Peppas was elected a member of the Institute of Medicine of the NAS. Having four faculty members elected to this assembly of world-class scholars is a stellar achievement for any campus in the nation.

Go to the Web sites of these faculty, Google their names, read their publications. You will find that statistics and scientific computation are critical to one and all.

  • Clean, Renewable, Sustainable Energy: Oil goes to $147 a barrel and then slides back. The need for energy shapes political campaigns, drives public policy, influences the economy and threatens national security. Global warming and the need for clean energy sources become topics of dinner conversations and the Nobel Prize.

Chemists, biologists, computer scientists, physicists, geologists, engineers, economists and public policy faculty across campus have turned attention to these matters. College-crossing programs have been built. Multi-million dollar grants and endowments have been secured. Data have been collected and archived.

Statistical techniques are employed and refined to mine these rich and otherwise overwhelming data sets. Resulting models provide mathematical and visual representations of uncertain, yet predictably patterned outcomes – from promising molecular structures, to wind corridors, to geological substrata, to financial markets. Students are trained to understand and explore these models and matters and thereby maximize their chances of productive careers in an uncertain labor market.

When it comes to what we know and what we do about clean, renewable, sustainable energy, the importance of statistics and scientific computation can hardly be over emphasized.

  • Social Policy: As Alan Greenspan repeatedly reminds us in “The Age of Turbulence,” the patterned uncertainties of econometric forecasts have come to dominate national and global economic forecasts and policy decisions. Simultaneous-equation models are used with uncertain success to explore supply and demand, to develop monetary policy and to examine the likely impact of such stimuli as a 25 basis point rise in the federal interest rate?

Polling agencies are now ubiquitous. Political strategists use multivariate statistical models to design winning campaigns and forecast election outcomes. The patterned uncertainties of consumer confidence, product preferences, life satisfaction, religious beliefs and practices, as well attitudes toward social issues such as abortion and capital punishment are now routinely mapped using estimated probabilities gleaned from stratified random samples.

Similarly employing the laws of probability and multivariate statistical models, demographers and healthcare officials investigate shifts in population growth, the etiology of infectious diseases, the allocation of health care costs and projections of life expectancies.

On this campus, admission directors use statistical models to predict the uncertain decisions of aspiring young adults. Similar models map the uncertainties of success in the classroom. Our nationally replicated UTeach program in math and science examines data to map and develop successful strategies to fill the pipeline with high quality teachers.

If you are interested in policy making and influencing the world in which we live, you will do well to become informed about statistical models of patterned uncertainties.

  • Hurricanes and Earthquakes: “Katrina,” “Ike” and “Sumatra Tsunami” bring images of devastating natural disasters quickly to mind. Likewise, if you grew up in California you would know precisely what the phrase “waiting for the big one” implies.

The ability to accurately predict earthquakes and their most likely timing and magnitude is yet to come. Pre-detection and location forecasting, however, are within reach. Both rely on patterned uncertainties and the probabilities emerging from statistical examination of geological data.

Dynamic statistical models, based on uncertain but well structured probabilities, help us anticipate the most likely paths and intensities of hurricanes. They produce projections based on global as well as more local data. The resulting, and now familiar, five-day projection cones are used to narrow the range of expected landfall and help minimize the loss of life and property damage.

Some models produce trajectories in a matter of minutes, using relatively small computers. Others, relying on dynamic global atmospheric, temperature, wind patterns and humidity can take hours to run on the world’s most advanced computers, whose computing power is measured in terms of a trillion Floating Operations per Second (TERAFLOPs), soon to be petaflops, and then on to exaflops.

Whether quick, static and relatively easy, or large, dynamic and complicated, statistical and scientific computation models are critical for those, which includes us all, interested in the prediction and pre-detection of natural disasters.

  • Just for Fun: Like many who spent their youth when baseball was the national pastime, I first gained profound respect for statistics by memorizing and calculating batting and earned-run averages, advancing in upper division calculations to slugging percentages. As my interests expanded, the significance of proportion free-throws-made, and more recently, shooting efficiency from three-point range loomed large.

Books are written on how team owners and coaches use these data in multivariate models to recruit talent. A section of the American Statistical Association is dedicated to sports with a range of sites for cycling, fencing, golf, hockey, horse racing, soccer, tennis and volleyball, to name just a few. There is now even a Journal of Quantitative Analysis in Sports. If you want a quick overview of this field, you might get copy of “Anthology of Statistics in Sports.”

Whatever your interests — hurricanes, neural networks, public policy, political campaigns, sports — you would do well to develop a deeper understanding of statistics and the fine points of scientific computation. Drop by the Division of Statistics and Scientific Computation in WCH 2.104. There is a whole world of patterned uncertainties waiting for you.