Robust Regression Using Data Partitioning and M-Estimation

We propose a new robust regression estimator using data partition technique and M estimation (DPM). The data partition technique is designed to define a small fixed number of subsets of the partitioned data set and to produce corresponding ordinary least square (OLS) fits in each subset, contrary to the resampling technique of existing robust estimators such as the least trimmed squares estimator. The proposed estimator shares a common strategy with the median ball algorithm estimator that is obtained from the OLS trial fits only on a fixed number of subsets of the data. We examine performance of the DPM estimator in the eleven challenging data sets and simulation studies. We also compare the DPM with the five commonly used robust estimators using empirical convergence rates relative to the OLS for clean data, robustness through mean squared error and bias, masking and swamping probabilities, the ability of detecting the known outliers, and the regression and affine equivariances.     

Automatic model identification using vector sample autocorrelation function

One of the weaknesses of the ESACF approach for model identification by Tsay and Tiao(1984) is the ambiguity which can be caused by the elements which are in triangle but marginally larger than two standard deviation values. To avoid this drawback, a vector sample autocorrelaton function (VSACF) is defined and an automatic model identification procedure using the VSACF is developed. We illustrate this approach with four examples.     

Asymptotic equivalence between two empirical mean residual life processes

There are two mean residual life estimates for right censored data. One is based on the Kaplan-Meier estimate, the other, based on the Susarla-Van Ryzin estimate for survival function. In this paper, we define the empirical mean residual life process for right censored data and show that the two empirical mean residual life processes based on the Kaplan-Meier and Susarla-Van Ryzin estimates are asymptotically equivalent uniformly on an interval under some conditions. Also we discuss the case which the asymptotic equivalence might fail.     

Modified Robust Second-Order Slope-Rotatable Designs

Generally it is very difficult to construct robust slope-rotatable designs along axial directions. Present paper focuses on modified second-order slope-rotatable designs (SOSRDs) with correlated errors. Modified robust second-order slope-rotatability conditions are derived for a general variance–covariance structure of errors. These conditions get simplified for intraclass correlation structure. A few robust second-order slope-rotatable designs (over all directions, or with equal maximum directional variance slope, or D-optimal slope) are examined with respect to modified robust slope-rotatability. It is observed that robust second-order slope-rotatable designs over all directions, or with equal maximum directional variance slope, or D-optimal slope are not generally modified robust second-order slope-rotatable designs.     

A measure and a graphical method for evaluating slope rotatability in response surface designs

A measure for evaluating slope rotatability over all directions in response surface designs, is proposed. This measure is used to form slope variance dispersion graph evaluating the overall slope rotatability and the slope estimation capability of an experimental design throughout the region of interest. This graph allows for an easy comparison of competing designs.     

A practical extension of the generalized estimating equation approach for longitudinal data

Liang and Zeger (1986) proposed an extension of generalized linear models to the analysis of longitudinal data. In their formulation, a common dispersion parameter assumption across observation times is required. However, this assumption is not expected to hold in most situations. Park (1993) proposed a simple extension of Liang and Zeger's formulation to allow for different dispersion parameters for each time point. The proposed model is easy to apply without heavy computations and useful to handle the cases when variations in over-dispersion over time exist. In this paper, we focus on evaluating the effect of additional dispersion parameters on the estimators of model parameters. Through a Monte Carlo simulation study, efficiency of Park's method is compared with the Liang and Zeger's method.     

Short-Term Prospective Effects of Impulsivity on Binge Drinking: Mediation by Positive and Negative Drinking Consequences

Abstract. Objective: Although the association of impulsivity with diverse alcohol outcomes has been documented, the mechanisms by which impulsivity predicts drinking over time remain to be fully characterized. The authors examined whether positive drinking consequences, but not negative drinking consequences, mediated the association between impulsivity and subsequent binge drinking, over and above prior binge drinking. Participants: Participants were 171 college students. Methods: Participants completed 2 online surveys with an average interval of 68 days between assessments at Time 1 (September to October 2012) and Time 2 (November to December 2012). Results: Path analysis showed that, among 5 facets of impulsivity, the effect of sensation seeking on subsequent binge drinking was completely mediated by prior positive consequences. No mediating effects of negative consequences were found. Conclusions: Prior experience of positive drinking consequences may serve as one of the risk pathways by which sensation seeking shapes binge drinking over time. Personalized intervention strategies may utilize information about students’ impulsivity facets to address their binge drinking and alcohol-related consequences.     

The rise of one-person households and their recent characteristics in Korea

"This paper explores the rise in the proportion of one-person households [in South Korea] which since 1960 was attributable largely to rural-to-urban migration of young people. The heavy migration of young people from rural to urban cities actually brought about the rural household division which contributed to the fall in household size, and contributed to the increase in one-person households of urban young singles, on the one hand, and the increase in one-person households of rural elderly widows, on the other. The rise in the aggregate propensity to live alone was also evident.... Although the rise in one-person households was evident and by 1990 about one out of every ten households was a one-person household, the findings on the recent characteristics of one-person households do not indicate any big change in the traditional family norms."     

The Evolution of Korean corporate HRD: launching,growing pains,and transforming

Enthusiasm and greater attention from the international HRD community to Korea has motivated us to conduct research on the evolution of Korean corporate HRD in order to explore where we are now and where we are heading. Research literature on the topic, however, was minimal. The purpose of this study was to examine the evolution of Korean corporate HRD through interviews with 18 HRD researchers and practitioners who have witnessed the field’s growth. In addition, we reviewed the literature and company archives of HRD activities. We identified three major historical milestones of Korean corporate HRD including: the launching period in the 1980s, the growing pains period in the 1990s, and the transforming period from the 2000s to the present. We addressed major trends and issues of Korean corporate HRD in an attempt to generate a future research agenda.