A non-parametric statistic for testing conditional heteroscedasticity for unobserved component models

When prediction intervals are constructed using unobserved component models (UCM), problems can arise due to the possible existence of components that may or may not be conditionally heteroscedastic. Accurate coverage depends on correctly identifying the source of the heteroscedasticity. Different proposals for testing heteroscedasticity have been applied to UCM; however, in most cases, these procedures are unable to identify the heteroscedastic component correctly. The main issue is that test statistics are affected by the presence of serial correlation, causing the distribution of the statistic under conditional homoscedasticity to remain unknown. We propose a nonparametric statistic for testing heteroscedasticity based on the well-known Wilcoxon''s rank statistic. We study the asymptotic validation of the statistic and examine bootstrap procedures for approximating its finite sample distribution. Simulation results show an improvement in the size of the homoscedasticity tests and a power that is clearly comparable with the best alternative in the literature. We also apply the test on real inflation data. Looking for the presence of a conditionally heteroscedastic effect on the error terms, we arrive at conclusions that almost all cases are different than those given by the alternative test statistics presented in the literature.     

Accurate estimation for extra-Poisson variability assuming random effect models

In this study, the components of extra-Poisson variability are estimated assuming random effect models under a Bayesian approach. A standard existing methodology to estimate extra-Poisson variability assumes a negative binomial distribution. The obtained results show that using the proposed random effect model it is possible to get more accurate estimates for the extra-Poisson variability components when compared to the use of a negative binomial distribution where it is possible to estimate only one component of extra-Poisson variability. Some illustrative examples are introduced considering real data sets.     

A comparison study on modeling of clustered and overdispersed count data for multiple comparisons

Data collected in various scientific fields are count data. One way to analyze such data is to compare the individual levels of the factor treatment using multiple comparisons. However, the measured individuals are often clustered – e.g. according to litter or rearing. This must be considered when estimating the parameters by a repeated measurement model. In addition, ignoring the overdispersion to which count data is prone leads to an increase of the type one error rate. We carry out simulation studies using several different data settings and compare different multiple contrast tests with parameter estimates from generalized estimation equations and generalized linear mixed models in order to observe coverage and rejection probabilities. We generate overdispersed, clustered count data in small samples as can be observed in many biological settings. We have found that the generalized estimation equations outperform generalized linear mixed models if the variance-sandwich estimator is correctly specified. Furthermore, generalized linear mixed models show problems with the convergence rate under certain data settings, but there are model implementations with lower implications exists. Finally, we use an example of genetic data to demonstrate the application of the multiple contrast test and the problems of ignoring strong overdispersion.     

Institutionalized collective action and the relationship between beliefs about environmental problems and environmental actions: A cross-national analysis

How do institutions affect the relationship between an individual's beliefs and their actions? Institutionalized strategies are routine ways of addressing problems that become taken-for-granted in a society. Environmental problems constitute a collective action problem in that personal consumption often conflicts with collective interests. I test whether beliefs about environmental problems have a different impact on a person's pro-environmental behaviors, depending on how addressing collective action problems is institutionalized in their society. In particular, I use level of welfare targeting as an observable, organizational difference among societies that reflects different institutionalized strategies for addressing a prominent collective action problem. I use multilevel models on data from the 2010 International Social Survey Programme (ISSP) and measures of welfare targeting from the Organization for Economic Cooperation and Development (OECD) to answer this question. I find that in societies where the institutionalized strategy for dealing with inequality is highly targeted, individuals' beliefs that these problems are important, real, and whether they can do something about them have a greater impact on their actions. The results suggest individuals generalize taken-for-granted strategies of assuring collective welfare to implement their individual beliefs about the environment, making institutional environments important moderators of the strength of the belief-action relationship.     

Changes in Parenting and Youth Adjustment Across the Military Deployment Cycle

This study examined how changes in at‐home parents' mental health and parenting practices related to changes in their children's adjustment throughout the course of a service members' military deployment. Participants included at‐home parents from 114 National Guard families who were interviewed at four different occasions across the deployment cycle. The results revealed changes across the deployment cycle among the following three indicators: parental warmth, depressive symptoms, and children's externalizing behaviors. Changes in parental warmth were associated with changes in children's adjustment. Overall, these findings indicate that during parental separation, at‐home parents' responses to children have important implications for children's adjustment.     

New evidence about effects of reproductive variables on child mortality in sub-Saharan Africa

There is still considerable uncertainty about how reproductive factors affect child mortality. This study, based on Demographic and Health Survey data from 28 countries in sub-Saharan Africa, shows that mortality is highest for firstborn children with very young mothers. Other children with young mothers, or of high birth order, also experience high mortality. Net of maternal age and birth order, a short preceding birth interval is associated with above average mortality. These patterns change, however, if time-invariant unobserved mother-level characteristics of importance for both mortality and fertility are controlled for in a multilevel–multiprocess model. Most importantly, there are smaller advantages associated with longer birth intervals and being older at first birth. The implications of alternative reproductive ‘strategies’ are discussed, taking into account that if the mother is older at birth, the child will also be born in a later calendar year, when mortality may be lower.     

On the use of quasi-likelihood for estimation in hidden-Markov random fields

This paper illustrates the use of quasi-likelihood methods of inference for hidden Markov random fields. These are simple to use and can be employed under circumstances where only the model form and its covariance structure are specified. In particular they can be used to derive the same estimating equations as the E-M algorithm or change of measure methods, which make full distributional assumptions.     

Univariate and multivariate models of positive and negative networks: Liking,disliking, and bully–victim relationships

Three relations between elementary school children were investigated: networks of general dislike and bullying were related to networks of general like. These were modeled using multivariate cross-sectional (statistical) network models. Exponential random graph models for a sample of 18 classrooms, numbering 393 students, were summarized using meta-analyses. Results showed (balanced) network structures with positive ties between those who were structurally equivalent in the negative network. Moreover, essential structural parameters for the univariate network structure of positive (general like) and negative (general dislike and bullying) tie networks were identified. Different structures emerged in positive and negative networks. The results provide a starting point for further theoretical and (multiplex) empirical research about negative ties and their interplay with positive ties.     

Apron strings of working mothers: Maternal employment and housework in cross-national perspective

This paper asks whether maternal employment has a lasting influence on the division of household labor for married women and men. Employing multi-level models with 2002 ISSP survey data for 31 countries, we test the lagged accommodation hypothesis that a long societal history of maternal employment contributes to more egalitarian household arrangements. Our results find that living in a country with a legacy of high maternal employment is positively associated with housework task-sharing, even controlling for the personal socialization experience of growing up with a mother who worked for pay. In formerly socialist countries, however, there is less gender parity in housework than predicted by the high historical level of maternal employment.     

Evolution of product quality in European,Japanese and US automotive firms: An exploratory longitudinal analysis

For years we have been hearing that US automobile manufacturers have been losing market share to their Japanese rivals who are reputed to make better quality vehicles. Most such reports are based on the initial quality surveys on new automobiles. In this paper we address two exploratory questions: (1) how does the quality of an automobile change with its age, and, (2) can firm level variables help explain differences quality. To answer these questions, we collected Consumer Reports’ reliability ratings on approximately 300 automobile models made by European, Japanese and US automotive firms during the 1998–2007; and approximately 240 models made by these firms over period of 2008–2015. For both periods we found that not only do automobiles made by Japanese firms have higher initial quality, but, as automobiles get older the difference in the product quality between Japanese versus European and US firms increases. We also found that the more generalist a European or US automobile firm, i.e., the wider is the firm׳s product offering in the marketplace, the lower its overall automobile quality during the 1998–2007 period. Conversely, Japanese generalist firms were found to have higher quality than specialist firms over the same period. The result is partly explained by the fact that Japanese firms have taken a different path to broadening their product variety – they have ensured a high level of quality of their initial offerings before entering newer market segments. The rate of reliability decline was found to be slower for all firms, and the differences in reliability across the 3 groups of firms were much less pronounced during the 2008–2015 period. This improvement may be as result of restructuring done by US automobile firms.