Main research problems in the training of researchers

Abstract

In this last chapter I will try to summarise the main arguments made in the preceding chapters and the points raised during the Tokyo IASE Round Table Conference debates. During a very productive week at the Institute of Statistical Mathematics, we discussed the issues and problems concerning the Training of Researchers in the Use of Statistics. As suggested by Jolliffe, this is an extensive topic that involves consideration of statistics itself, the interaction between statisticians and researchers, and even the training of statisticians. My aim in writing this summary is to provide the reader with the view that emerged from our joint reflections about this complex issue and its many facets. The conclusions below are organised around the research questions initially set out in the Discussion Document, which is reproduced in the first chapter of the book, so that the extent to which those questions were taken into account in the different contributions can be more easily perceived. All the authors highlighted the relevance of improving researchers' statistical training and of making statisticians better acquainted with other research fields for ensuring an optimal communication between statisticians and researchers. However, it was also made clear from the arguments throughout the book that there are no easy or definitive answers to the questions raised. In addition to the complexity of the topic, little research has been done until now on the teaching of statistics to postgraduates and on the difficulties of learning advanced statistical topics. Even when some of the papers present examples of useful research and innovative teaching experiences, the Conference participants felt that there is an urgent need to increase research in this area. This chapter then presents not a list of final agreed solutions to the problems of training of researchers, but our view about the main areas in which further research on this field is needed. Some recommendations about ways in which the teaching of statistics might contribute to increasing statistical knowledge among researchers and to better use of statistics in experimental research are also made.