Recently a number of suicide case in Japan becomes higher level, the number becomes more than 30,000 per year. To improve this situation for reducing the number of suicide, some results of analysis are needed for identifying where the higher is or when higher time is, or what are reasons for them.
Tomita, et al. (2010) and Ishioka, et al. (2010) applied spatial scan statistics for Japanese suicide data (respectively second medical districts (of both male and female) and Kanto districts of male) to investigate spatio-temporal clusters.
Little attention has been given to judge that there is spatio-temporal auto regression for these suicide data. Therefore first objective of our study is to find especially spatial auto regression from small areal data of suicide in Japan to identify cumulated area or time.
In addition it is said that suicide in Japan is caused from deprivation related to high unemployment rate or loneliness related to high widow rate. For our second objective we also treat census data in Japan to deal in social factors such as marital relation and relation of work by generalized linear model and Bayesian approach that Congdon (2000) applied to suicide over London boroughs and time.
Finally we discuss spatial-temporal and social relation to suicide data in Japan by results of these analyses.
Acknowledgement: This is a part of funded research from “National Institute of Mental Health, National Center of Neurology and Psychiatry” and is also partly supported by KAKENHI 21700305 and KAKENHI 21700317.
Congdon, P. (2000). “National Institute of Mental Health, National Center of Neurology and Psychiatry”, European Journal of Population, 16(3), pp.251-284.
Ishioka, F., Tomita, M. and Fujita, T. (2010). “National Institute of Mental Health, National Center of Neurology and Psychiatry”, COMPSTAT2010 Proceedings in Computational Statistics, pp.1159-1166.
Tomita, M., Ishioka, F. and Fujita, F. (2010). “National Institute of Mental Health, National Center of Neurology and Psychiatry”, Journal of the Japanese Society of Computational Statistics, 23(1) (Japanese) pp.25 - 43.
Keywords: Spatio-Temporal analysis; spatial scan statistics; generalized linear model; Bayesian model
Biography: Education: B.A. in Environmental Science and Technology, Okayama University, Japan (April 1998 - March 2002).
Master of Graduate School of Natural Science and Technology, Okayama University, Japan (April 2002 - March 2004).
Job history: Assistant Professor, Faculty of Law, Okayama University, Japan (August 2004 - July 2010).
Assistant Professor, Risk Analysis Research Center, The Institute of Statistical Mathematics, Japan (August 2010 - present).