Communication of statistcal concepts: examples in medical collaboration

Abstract

While effective communication of statistical concepts is important for the enthusiastic adoption of these concepts by collaborators, statisticians are not necessarily trained in the process of communication with collaborators in other substantive fields. It is proposed that increased attention be paid to pedagogical techniques for communicating to our non-statistical colleagues what statisticians have to offer to the design and analysis aspects of a collaborative effort. One approach is to offer examples relevant to our colleagues’ fields when we explain statistical ideas. This paper provides several such examples from the field of neurology, focusing on the issue of sample selection bias and prospective study designs.

Keywords:

• Communication
• sample selection bias
• prospective studies
• outcome assessment
• diagnosis
• prevalence

^∗Jonas H. Ellenberg is Vice President and Senior Biostatistician, Westat Rockville, Maryland 20850. This article is the American Statistical Association’s President’s Address delivered at the 50th Anniversary Conference of the Department of Statistics at Virginia Polytechnic Institute and State University in Blacksburg, Virginia, on August 13, 1999.

^∗Jonas H. Ellenberg is Vice President and Senior Biostatistician, Westat Rockville, Maryland 20850. This article is the American Statistical Association’s President’s Address delivered at the 50th Anniversary Conference of the Department of Statistics at Virginia Polytechnic Institute and State University in Blacksburg, Virginia, on August 13, 1999.

Notes

^∗Jonas H. Ellenberg is Vice President and Senior Biostatistician, Westat Rockville, Maryland 20850. This article is the American Statistical Association’s President’s Address delivered at the 50th Anniversary Conference of the Department of Statistics at Virginia Polytechnic Institute and State University in Blacksburg, Virginia, on August 13, 1999.