Ricardo Piper Segurado (University College Dublin)

will speak on

Notes from the biostatistical front line

Time: 3:00PM
Date: Thu 31st October 2019
Location: Seminar Room SCN 1.25 [map]

Abstract: It’s well known that the practice of data analysis in research in the biological and health sciences suffers from frequent mis-use of statistical procedures, and tends to lag behind what might be best practice to an academic statistician, in some cases by decades. Why this occurs may be a mixture of many factors related to training of teachers, teaching methods, historical inertia, what you might call structural incentives against correction, as well as various difficulties in the implementation of solutions. I will present some case studies that explore technical errors, interpersonal misunderstandings, as well as indicating where structural factors can have a crucial role. I will frame these in the light of the recent 'reproducibility crisis', and attempt to provoke a broad discussion around the role of statisticians in biomedical research.

(This talk is part of the Statistics and Actuarial Science series.)

PDF notice

Return to all seminars