Quantile Regression Methods for Longitudinal Data with Drop-outs: Application to CD4 Cell Counts of Patients Infected with the Human Immunodeficiency Virus

Patients infected with the human immunodeficiency virus (HIV) generally experience a decline in their CD4 cell count (a count of certain white blood cells). We describe the use of quantile regression methods to analyse longitudinal data on CD4 cell counts from 1300 patients who participated in clinical trials that compared two therapeutic treatments: zidovudine and didanosine. It is of scientific interest to determine any treatment differences in the CD4 cell counts over a short treatment period. However, the analysis of the CD4 data is complicated by drop-outs: patients with lower CD4 cell counts at the base-line appear more likely to drop out at later measurement occasions. Motivated by this example, we describe the use of `weighted' estimating equations in quantile regression models for longitudinal data with drop-outs. In particular, the conventional estimating equations for the quantile regression parameters are weighted inversely proportionally to the probability of drop-out. This approach requires the process generating the missing data to be estimable but makes no assumptions about the distribution of the responses other than those imposed by the quantile regression model. This method yields consistent estimates of the quantile regression parameters provided that the model for drop-out has been correctly specified. The methodology proposed is applied to the CD4 cell count data and the results are compared with those obtained from an `unweighted' analysis. These results demonstrate how an analysis that fails to account for drop-outs can mislead.     

An analysis of variance type test for comparing clusters of DNA sequences based on randomization test methodologies

SUMMARY A method for comparing groupings of DNA sequences is presented, which utilizes randomization test methods to assign significance levels to a test statistic defined in terms of the Hamming distance between two sequences. The method, which is intuitively motivated by the analysis of variance procedure, partitions the variation caused by differences between clusters from the variation attributable to differences at random base pair locations within clusers. Implementation issues are discussed, and an example of the application of the method is provided.     

Effects of Race and Sex on Work Group Cooperation Vs. Competition

This study examines whether differences in individualism-collectivism by race and sex result in differences in cooperation vs. competition. Expanding upon the earlier work of Cox et al., individual and group-level comparisons are made by race and sex. Exclusively examining race-based effects, Cox et al. had suggested that blacks display greater cooperation than do whites. Our findings concur; however, we are able to demonstrate this using a more complete group-level comparison. Given literature suggesting sex to be as relevant as race when administering work force diversity, we further expanded upon the earlier study by testing for such effects. While some previous research has suggested a tendency for women to behave more cooperatively than do men, our findings show that this is not necessarily the case. Specifically, this study demonstrates that sex decreases in importance, when controlling for race. Applications and suggestions for future research are presented.     

ARE AUTOMOBILE SAFETY REGULATIONS WORTH THE PRICE: EVIDENCE FROM USED CAR MARKETS

The value consumers place on automobile safety regulations can be determined by examining the instantaneous effects changes in these regulations have on used vehicle prices. If consumers value safety regulations more than their cost, used vehicle prices decrease in response to a change in safety regulations. If consumers place at least some value on the safety induced by safety regulations, a change in emission control regulations will have a greater positive impact on used automobile prices than a change in safety regulations. Empirical results strongly support the latter case and, surprisingly, offer some support for the former case. (JEL L51, L62)     

Eigenvalue—Free methods for computing the distribution of a quadratic form in normal variables

The traditional method for computing the distribution of a quadratic form in normal variables, Pr(v'Qv相似文献   

Shifting global supply networks and fast fashion: made in Turkey for Marks & Spencer

In the late 1980s and early 1990s, it was noted that retailers in Britain had started providing increased variety and fashionability to their customers, had added mid‐season purchasing to their previous two‐season calendars, and that a high fashion and low price ‘throwaway market’ had appeared on the retail scene. Since then mid‐season purchasing has evolved into purchasing throughout the year; and the ‘throwaway market’ (now called fast fashion) has become almost the norm. Here we revisit one of those British retailers (Marks & Spencer) together with its Turkish suppliers and observe a trend towards the diffusion of design capabilities to suppliers that is more widespread than is suggested in the literature. We also consider the question of how most appropriately to conceptualize the nature of these retailer‐ supplier relations in today's circumstances. We especially look into the manner in which these relations are reflected in price negotiations, the eventual distribution of the value‐added, and the nature of everyday interactions such as accreditation and reclamation practices. We conclude that even though Turkish suppliers seem to be successfully upgrading into design–a high value‐added activity–the question of whether or not this development has entailed a renegotiation of power between retailer and supplier remains a complicated one.     

Delayed Adulthood,Delayed Desistance? Trends in the Age Distribution of Problem Behaviors

As the transition to adulthood becomes more protracted and less orderly, fewer young people occupy adult roles and experience the social control associated with these roles. One might therefore expect behaviors associated with the teenage years to spill over into older age groups, reflecting postponed entrance into full social adulthood. We test this hypothesis by examining trends over time in the age distribution of crime, substance use, and violent death. We find little evidence that behaviors typical of adolescence are moving upward to older ages. Although the achievement of adult roles is being pushed to older ages, this stretching of the transition to adulthood is not reflected in the observed patterns of substance use, violent death, and arrests.     

NONPARAMETRIC AUTOCOVARIANCE FUNCTION ESTIMATION

Nonparametric estimators of autocovariance functions for non-stationary time series are developed. The estimators are based on straightforward nonparametric mean function estimation ideas and allow use of any linear smoother (e.g. smoothing spline, local polynomial). The paper studies the properties of the estimators, and illustrates their usefulness through application to some meteorological and seismic time series.