Every few days, we will be publishing layman’s abstracts of new articles from our prestigious portfolio of journals in statistics. The aim is to highlight the latest research to a broader audience in an accessible format.
The article featured today is from the Canadian Journal of Statistics and the full article, published in issue 47.3, is available to read online here.
Lin, H., Yang, B., Zhou, L., Yip, P. S., Chen, Y. and Liang, H. (2019), Global kernel estimator and test of varying‐coefficient autoregressive model. Can J Statistics, 47: 487-519. doi: 10.1002/cjs.11510
Taiwan has experienced a significant increase in suicide rates in the past two decades. It was as low as 6.2 per 100,000 (persons) in 1993 and reached a historical high of 19.3 per 100,000 in 2006. Suicide has become an urgent and important public health problem in Taiwan. The widespread media reporting of suicide events has been thought to be responsible for triggering copycat suicides and there was a mutual causation between suicide news reporting and suicide incidence; namely, the impact of media reporting on actual suicides is multiplicative and interactive. Although much research has explored the impact of the media coverage of suicide incidence, the current evidence for understanding the copycat suicide effect is still indirect and less clear.
This study aims to explore whether and how reporting of suicides affects suicide incidence. Particularly, this study proposed a varying-coefficient autoregressive model to address the following three specific questions: (i) How do the previous media coverage and the previous suicide rate affect the current suicide rate? (ii) Will the effect of media coverage diminish over time? If so, when and how do the previous media coverage and previous suicides become unrelated to current suicide incidence? (iii) Is the effect of previous suicides on current suicides amplified if they are fully reported in the media? If so, how is the effect amplified? This analysis allows nonparametric interactions and autoregression between the reporting intensity and suicide incidence. The results suggest that the effect of the previous suicide is a fold-linear effect amplified by the reporting on the next day, and the effect of the previous suicides and the reporting in the media does not diminish over time. The results provides empirical evidence to demonstrate the mutual causative relationship between media reporting and suicide incidence.