On Future Directions in Statistical Methodologies – Some Speculations


  • This paper is based on the SJS Lecture given at the NORDSTAT Meeting in Voss Norway, June 2010.

Elja Arjas, Department of Mathematics and Statistics, P.O. Box 68, FI-00014 University of Helsinki, Finland.
E-mail: elja.arjas@helsinki.fi


Abstract.  We, as statisticians, are living in interesting times. New scientifically significant questions are waiting for our contributions, new data accumulate at a fast rate, and the rapid increase of computing power gives us unprecedented opportunities to meet these challenges. Yet, many members of our community are still turning the old wheel as if nothing dramatic had happened. There are ideas, methods and techniques which are commonly used but outdated and should be replaced by new ones. Can we expect to see, as has been suggested, a consolidation of statistical methodologies towards a new synthesis, or is perhaps an even wider separation and greater divergence the more likely scenario? In this talk these issues are discussed, and some conjectures and suggestions are made.