Los Alamos National Laboratory
What is a statistician?
The stereotype of a statistician tabulating and summarizing masses of numbers fails to capture the diversity and creativity of statistical work. A statistician does collect and analyze data, but there are many aspects of this work, and as a statistician you will be involved in all of them. You will design sample surveys and laboratory experiments to maximize the information gained within time and budget constraints. You will modify standard analytical techniques so that they can be applied to the available data. You will also spend time educating students and colleagues about what statistics can and cannot do and learning from those colleagues about their problems.
Statisticians can be classified as “applied” or “mathematical,” although those with advanced degrees find that they wear either hat as the occasion requires. An applied statistician primarily works at solving problems for and with clients from other disciplines, frequently as a member of a multidisciplinary team. A mathematical statistician evaluates existing statistical techniques, devises new ones, and explores the underlying mathematical theory, often within the statistics or mathematics department of a university or other research institution.
What makes a good statistician?
The range of study areas in which statistics can be applied is enormous, and good statisticians take an interest in a broad spectrum of scientific and sociological problems. While they usually specialize to some extent, for example in applications related to medicine, engineering, or economics, they are versatile and enjoy learning about new areas of application throughout their lifetimes.
A critical skill is the ability to extract the important features of a problem from a mass of facts. A statistician must be at ease with mathematical concepts and must be able to formulate new problems in mathematical terms. In the last two decades, high-speed computers and new data-visualization techniques have revolutionized statistics, so it is essential to know how to use computers effectively.
Although they may prefer working with “things” rather than people, statisticians must be able to communicate with people outside their field as well as with their statistical colleagues. They must be both willing and able to go more than halfway to understand the problems and constraints of the researchers and other clients with whom they work, and to explain statistical conclusions to those with little or no statistical training.
What is life as a statistician like?
The long-term rewards come both from being part of the discovery process and from helping managers, teams, and researchers to make sound, data-based decisions. On a day-to-day basis, statisticians take pleasure in revealing the structure underlying a set of data and in using logic and mathematics to solve real-world problems. Like most jobs, statistics also requires a fair amount of plugging along at more mundane tasks, debugging computer programs, or proofreading reports. Perhaps the most frustrating aspect is the multiple demands on your time, keeping several projects going at once, and sometimes feeling unable to do justice to any of them.
If you are associated with a university, you will be teaching and helping students. As a statistician in any organization, you will spend time both on consulting on quick problems and learning in depth about new ones. You will need to review current journals both to find the right technique for a new problem and to keep abreast of new developments. You will probably do some traveling to meetings and short courses in the course of your consulting work.
How do I become a statistician?
A few universities offer a bachelor’s degree in statistics, but many statisticians obtain their bachelor’s degree in another field, such as a natural science, economics or sociology, mathematics, or computer science. With any degree that includes a strong background in mathematics (calculus, linear algebra, one or two courses in probability and statistics, and some computer science), you can work with other statisticians, conducting surveys and running standard data analyses.
Most statisticians complete a master’s degree, which equips them to work independently on applied statistical problems. Here the training is more specialized, including several courses on statistical theory and methods as well as valuable consulting experience under the guidance of a faculty member. It is important to take courses in one or more areas where you might apply statistics (e.g., in the natural or social sciences) and to learn to use the computer creatively as well as being familiar with the commercially available statistical software. As part of your general background, you might also take courses in technical writing and in the history and philosophy of science.
To become a mathematical statistician, teach at a university, or obtain a job in which you devote part of your time to research, you will probably need a Ph.D. degree. This involves further course work in specialized areas of statistics as well as writing a dissertation that represents an original contribution to the field of statistics.
What/where are the jobs?
The demand for statisticians continues to be strong. Industry employs about two thirds of all statisticians in many kinds of work: pharmaceutical research, quality control and reliability engineering, development of agricultural products, marketing and forecasting, and dozens of other areas. Government is concerned about demography, labor force surveys, natural resource estimation, and environmental monitoring, among other areas, and currently employs 10% to 15% of working statisticians. Established individuals with advanced degrees occasionally go into business for themselves, consulting for industry and government. Universities and colleges need statisticians with advanced degrees who will teach and collaborate with students and faculty in other fields.