First, we would like to express our sincere gratitude to all the authors, supporters, editorial board members of the journal Statistical Theory and Related Fields.
In this second issue of 2018, three types of articles are published, (1) a review paper on nonignorable missing data problems followed by five papers of discussions and an author’s rejoinder; (2) six original articles on statistical methodology and applications; and (3) a new created Statistics Column containing three articles related to an international statistics conference held in June 2018.
The review paper by Tang and Ju provides an overview of statistical methodology developed in recent year for problems with nonignorable missing data. Nonignorable missing data are frequently encountered in all fields of statistical application, and methodology development for handling nonignorable missing data is challenge because the missing data mechanism depends on unobserved data or other quantities. This article reviews recent advances in estimation, influence analysis, and model selection, with tools such as semiparametric modelling, empirical likelihood, case-deletion method and local influence analysis, information type criterion and penalisation for model selection, and Bayesian methods. This article is followed by five papers of discussions, each itself is a short communication paper in this important research field. The paper by Fang and Ni considers variable screening when covariates have missing data. The paper by Wang studies longitudinal data with nonignorable dropout. Morikawa and Kim in their paper propose a new idea in semiparametric estimation, which may lead to a more efficient estimator of the parameter in nonignorable missing data mechanism. The paper by Shao contains some new ideas of estimating the missing data mechanism which is one way to handle nonignorable missing data. On the other hand, the paper by Zhao discusses the other way of solving the problem without estimating the missing data mechanism. All these papers of discussions, together with the rejoinder by Tang and Ju, add more extensive results in the area of nonignorable missing data.
The next six papers are articles in six diversified areas of statistical interests. The paper by Vovan proposes a Bayes method in classification. The author establishes an algorithm using a Matlab procedure and applies it to the problem of evaluating ability of customers to pay debts at banks in Viet Nam. The article by Slud, Vonta, and Kagan studies combining estimators of a common parameter across samples, a problem of great interest in areas such as meta analysis, since nowadays datasets from different trials are available for more efficient estimation and inference. Xiong and Xu in their paper consider the problem of jointly estimating marginal quantiles of a multivariate distribution. They propose efficient procedures incorporating the correlation structure of multivariate distribution. In the paper by Liu, Shi, Bai, and Liu, the best linear unbiased estimators for coherent system dynamic stress-strength reliability, and the best linear unbiased predictors for the Type-II censored coherent system failure times are constructed. The authors provide exact expressions of the estimators and their variances and covariances. The paper by Chen, Xie, and Shao investigates sufficient dimension reduction for covariates in a semiparametric pseudo likelihood approach for data with nonignorable nonresponse. In the paper by Bourguignon and Vasconcellos, the authors use ranks and signs to deal with additive outliers in Poisson integer-valued autoregressive models.
The last part of this issue is a Statistics Column containing three articles, created for the international conference with title Small Area Estimation and Other Topics of Current Interest in Surveys, Official Statistics, and General Statistics, held between June 16-18, 2018 at East China Normal University, Shanghai, China. This conference is also a celebration of Professor Danny Pfeffermann’s 75th birthday. The first article by Tan and Ni summarises the three keynote lectures presented at the conference by professors Berger, Rao, Ghosh, which mainly review the previous studies and present the pioneering results covering Bayesian statistics, small area estimation, and shrinkage priors. The second article by Ni and Tan is about an interview with professor Danny Pfeffermann. The last article is professor Pfeffermann’s speech on ‘how I become a statistician’ at a birthday dinner during the conference.
Finally, we welcome submissions of high-quality interesting and diversified articles from statisticians and scientists analyzing various types of data.