The Institute runs a unique system to promote collaborative research activities between statisticians and scientists in related fields such as social sciences, the humanities, medical sciences and engineering. The system was initiated in 1985 with a special intention, which has much to do with the past experience of the Institute. Since the very beginning of the history of the Institute, one of the basic principles has been to attach a great importance to applications. The principle came from appreciating that innovative methodologies and theories of statistics are frequently developed in an effort to solve real problems.
In the past decades the Institute has maintained research collaborations between universities, government offices, private companies and various organizations. During this time, many useful works, both in theory and application, have been produced. This tradition of open collaboration with scientists outside the Institute has created a progressive and liberal academic atmosphere which, we believe, has contributed to developing new interdisciplinary research fields in related sciences.
The collaborative research activity was maintained through various research fields with various levels and types of collaboration, long before the Institute was reorganized into an inter-university research institute. Many remarkable results have been produced through collaborative research in the last decades. To our regret, however, when joint work is organized by researchers at the individual level, the fruit of the collaborative research tends to be received, by the general public, as a successful contribution to the science where the solved problems arose, even when our statisticians played the most essential role. Obviously this tendency comes from the inherently abstract nature of statistics. The statistician's contribution, although essential, is not as easy to explain to the public at large as explaining the problem itself in applied science. Accordingly, it seemed that the value and the raison d'être of the statisticians and the Institute was not appreciated as much as other scientists and research institutes in the applied sciences.
Our present collaborative research system was initiated on the basis of two understandings. Firstly, this kind of collaborative research activity is beneficial to both statistics and other related sciences. Secondly, the statisticians working in such circumstances need recognition, support and encouragement. We hope that the present system will play a role similar to the one hospitals play in the medical sciences. Without constant stimuli from patients in the hospital little development in medical sciences would be expected.
Since 1985 the system has been run by the Co-operative Research Committee, half of whose members are scientists from outside the Institute. Collaborative research projects between statisticians and scientists in related scientific fields are called for each year. More than a hundred projects in applied sciences and statistics are supported each year (see the figure below). Projects are classified into two categories; methodologies and applications. Projects in methodologies are further classified into four categories; basic theory, computation, time series and sampling. Projects in applications are classified into 5 categories; physical and engineering sciences, space and earth sciences, bio-medical sciences, human and social sciences, and lastly, environmental sciences.
The system of collaboration is open to projects that are to be planned and accomplished through international cooperation.
In 1998, in hope of enlarging the area of collaboration, the Institute relaxed a condition of application for projects which had stipulated that at least one member of the research project should belong to the Institute.
Number of collaborative research projects