We present a novel approach to test for heteroscedasticity of a non-stationary time series that is based on Gini's mean difference of logarithmic local sample variances. In order to analyse the large sample behaviour of our test statistic, we establish new limit theorems for U-statistics of dependent triangular arrays. We derive the asymptotic distribution of the test statistic under the null hypothesis of a constant variance and show that the test is consistent against a large class of alternatives, including multiple structural breaks in the variance. Our test is applicable even in the case of non-stationary processes, assuming a locally stationary mean function. The performance of the test and its comparatively low computation time are illustrated in an extensive simulation study. As an application, we analyse data from civil engineering, monitoring crack widths in concrete bridge surfaces.

No abstract

With high urbanisation rates, cities in sub-Saharan Africa are contending with food insecurity. Urban studies scholars have approached the issue mainly from the perspective of food deserts. We adapt Sen’s ‘resource bundles’ and Watts and Bohles’s ‘space of vulnerability’ concepts to examine food insecurity as a function of both tangible and intangible resources. Moreover, we also interrogate the role of kin in strengthening safety nets for the urban poor. Drawing on a data set of 462 single mothers in a slum in Nairobi, Kenya, we find that (1) bundles comes in four types; (2) bundles with high levels of all resources buffer against food insecurity as do (3) bundles weighted with high levels of wealth and social standing; and (4) kin enhance the protective effect of bundles only for two types. These findings should direct urban poverty researchers to consider the compounding effect of resources in the reproduction of poverty and social inequality and encourage policy makers to focus on both vulnerability and resilience in designing interventions to ensure food security.

Background and Objectives Graft failure (GF) after cord blood transplant (CBT) has decreased with improved supportive care and cord selection strategies. We aimed to evaluate cord blood selection and factors associated with retransplantation on the incidence of GF, determine risk factors for GF including host antibodies to Kell antigen and evaluate survival after GF. Materials and Methods We retrospectively reviewed 84 patients who underwent CBT at the University of Oklahoma between 2000 and 2016 and compared outcomes in patients with/without engraftment by Day 28. The nonengraftment cohort was further divided into patients who underwent retransplantation. Kaplan–Meier curves with log‐rank tests were calculated to assess the association between mortality and engraftment. Results Engraftment following CBT was high at 81%, with 52% engrafting by Day 28 and an additional 29% engrafting by a median of 36 days. Retransplantation led to 88% engraftment at a median of 53 days. Overall, 75% of the 40 patients who did not engraft by Day 28 died. Female sex and total nucleated cell count < 3.5/kg were significantly associated with lack of engraftment and higher mortality. Antibodies to Kell fetal antigen were not identified. Retransplantation by Day 28 for primary GF conferred a survival advantage. Conclusion This study demonstrates that failure to engraft by 28 days was associated with increased mortality, and risk was mitigated with early retransplantation. Female sex and low total cell dose were associated with increased mortality. Early identification of GF coupled with early retransplantation can reduce mortality in CBT.

We propose a new asymptotic test to assess the stationarity of a time series' mean that is applicable in the presence of both heteroscedasticity and short-range dependence. Our test statistic is composed of Gini's mean difference of local sample means. To analyse its asymptotic behaviour, we develop new limit theory for U-statistics of strongly mixing triangular arrays under non-stationarity. Most importantly, we show asymptotic normality of the test statistic under the hypothesis of a constant mean and prove the test's consistency against a very general class of alternatives, including both smooth and abrupt changes in the mean. We propose estimators for all parameters involved, including an adapted subsampling estimator for the long run variance, and show their consistency. Our procedure is practically evaluated in an extensive simulation study and in two data examples.