A resampling approach to estimate variance components of multilevel models

In a multilevel model for complex survey data, the weight‐inflated estimators of variance components can be biased. We propose a resampling method to correct this bias. The performance of the bias corrected estimators is studied through simulations using populations generated from a simple random effects model. The simulations show that, without lowering the precision, the proposed procedure can reduce the bias of the estimators, especially for designs that are both informative and have small cluster sizes. Application of these resampling procedures to data from an artificial workplace survey provides further evidence for the empirical value of this method. The Canadian Journal of Statistics 40: 150–171; 2012 © 2012 Statistical Society of Canada     

Concomitants of k-record values arising from Morgenstern family of distributions and their applications in parameter estimation

In this paper, we have obtained the marginal and joint distributions of concomitants of k-record values for the Morgenstern family of distributions (MFD) and hence obtained the moments and product moments of concomitants of k-record values. Applying this results we have derived the best linear unbiased estimators of some parameters involved in Morgenstern type bivariate logistic distribution which belongs to MFD based on concomitants of k-record values.     

Gambling as a function of gender and Sensation Seeking

This study examined the relation of gender and Sensation Seeking (SS) Scale scores to undergraduates' ratings of their past/present and expected future participation in four leisure activities that usually involve betting money. Subjects completed McKeachnie's (1975) Leisure Activities Blank (LAB), which included the gambling items, and Zuckerman's (1979) SS Scale. Men reported significantly more past/present leisure gambling than women; gender differences were insignificant in ratings of expected future gambling. No significant differences in reported past/present gambling related to any SS scale score. However, expected future gambling ratings were associated with scores on two SS subscales: Disinhibition (DIS) and Boredom Susceptibility (BS). Both male and female subjects with high DIS scores rated the frequency of expected future gambling higher than did subjects with low DIS scores. High BS scores were associated with high ratings of expected future gambling in the data for women subjects only. These results suggest that personality factors, among them SS, are more influential than early experience or sex-role socialization in determining a young adult's interest in gambling.     

Kin Effects on Black-White Account and Home Ownership

This article combines the literature on kin networks and racial disparities in asset ownership. Specifically, we examine the effects of kin characteristics—sibling poverty and parental poverty, education, and occupation—on financial account ownership and home ownership. We find that kin matter for these outcomes. Having a poor sibling and coming from a poor family are negatively associated with account and home ownership while mother's education has a positive effect. Separate analyses by race suggest that kin characteristics matter for both Blacks and Whites for account ownership, but for home ownership they are significant for Whites only. Racial differences in kin characteristics account for over half of the racial gap in account ownership, but are not important for understanding the racial gap in home ownership. The significant effects of extended family characteristics on socioeconomic well-being make a case for the inclusion of kin variables in the growing literature on wealth disparities among Blacks and Whites.