Journal of Applied Statistics Best Paper Prize 2017
Journal of Applied Statistics
About the prize
The 2017 Journal of Applied Statistics Best Paper Prize had been decided.
The award goes to Connie Stewart for her paper: ‘An approach to measure distance between compositional diet estimates containing essential zeros’
This award was decided by the Editor-in-Chief with the support of the Associate Editors, their comments on the winning paper are as follows: “…Excessive zeros are commonly encountered in many statistical applications such as dietary qualities (zeros due to non-consumption of certain foods), physical activities (zero if no activity), and insurance claims (zero if no claims). Statistical Analysis results on data that have excessive zeros can be misleading if these zero values are not properly handled during the analysis process”.
In this paper the author compared three measures of distance for compositional data capable of handling zeros, and showed that some well-accepted principles of compositional data analysis may not be satisfied, but the property of sub-compositional coherence may be approximately satisfied. From simulation results and an application to real data, the author recommended the chi-square measure of distance.
Since its publication, the paper has been viewed as a good reference for analysing dietary data from episodically consumed foods in a few nutritional studies conducted at the National Institutes of Health. The paper’s potential practical influence is predicted to be high and the paper demonstrated how a good application of statistical methodology can concretely contribute to the society at large.
You can read this winning paper, as well as the highly commended finalists, on this page.
Journal of Applied Statistics Best Paper Prize promotion
The Journal of Applied Statistics Best Paper Prize is awarded annually, as decided by the Editor-in-Chief with the support of the Associate Editors.
The winning article receives a £500 prize, and their paper will be made free to view for the following year. All articles published in the Journal are automatically included for consideration.
|2017-Highly Commended||Chen Peng, Maochao Xu, Shouhuai Xu & Taizhong Hu||Modeling and predicting extreme cyber attack rates via marked point processes||44||14|
|2017-Highly Commended||Daniel McNeish||Missing data methods for arbitrary missingness with small samples||44||1|