Research in statistics
Statistics is a method of science that is driven by an ever increasing need to collect data and draw conclusions from them. The subject has its roots in probability theory and mathematics and it covers all steps from collection to processing, analysis, interpretation and presentation of data.
Developments in computer technology have brought the subject of statistics into a new phase. Large, complex data sets are created at a rapid pace in research as well as in business and management. These form an invaluable source of information and knowledge. But new data sets and questions often require new statistical methodology. At the same time, increased computing power provides new possibilities for the development of effective tools.
Statistical methods are currently used in most empirical disciplines. Here are some examples of issues that require statistical analysis:
- Have extreme weather events become more common in recent years? Have the measures to reduce eutrophication of the Baltic Sea had an impact? How can we track the state of the environment when the sampling and measurement procedures are changed?
- What is the optimal level of the Riksbank's repo rate? How do you create a share portfolio with maximum return for a given level of risk? What factors determine if a company goes bankrupt? Which starting price in an auction gives the highest profit for the seller?
- How effectively can oil spills in the ocean be detected by satellite images? Does face recognition really work in practice? What is the lowest dose of radiation with which X-ray images of sufficient quality can be obtained?
- How important is the conformity in DNA from a blood trail from a crime scene and a saliva sample from a suspect? Is it sufficient to chemically examine just five pills in a large seizure in order to claim that most of the seizure consists of illegal pills?
The subject of statistics at Linköping University focuses on statistical methodology from an applied perspective. The work means that new statistical methodology is designed in consistency with the empirical problems, but also that existing models and methods are introduced in new contexts. The research often involves a large number of computer calculations.
The department's research is centred around the following research areas:
- Bayesian statistics
- Computer-intensive methods
- Forensic statistics
- Statistical genetics
Last updated: 2013-08-08