Fast,closed-form,and efficient estimators for hierarchical models with AR(1) covariance and unequal cluster sizes

This article is concerned with statistically and computationally efficient estimation in a hierarchical data setting with unequal cluster sizes and an AR(1) covariance structure. Maximum likelihood estimation for AR(1) requires numerical iteration when cluster sizes are unequal. A near optimal non-iterative procedure is proposed. Pseudo-likelihood and split-sample methods are used, resulting in computing weights to combine cluster size specific parameter estimates. Results show that the method is statistically nearly as efficient as maximum likelihood, but shows great savings in computation time.     

GMM inference in spatial autoregressive models

In this study, we investigate the finite sample properties of the optimal generalized method of moments estimator (OGMME) for a spatial econometric model with a first-order spatial autoregressive process in the dependent variable and the disturbance term (for short SARAR(1, 1)). We show that the estimated asymptotic standard errors for spatial autoregressive parameters can be substantially smaller than their empirical counterparts. Hence, we extend the finite sample variance correction methodology of Windmeijer (2005 Windmeijer, F. (2005). A finite sample correction for the variance of linear efficient two-step GMM estimators. Journal of Econometrics 126(1):25–51.[Crossref], [Web of Science ®] , [Google Scholar]) to the OGMME for the SARAR(1, 1) model. Results from simulation studies indicate that the correction method improves the variance estimates in small samples and leads to more accurate inference for the spatial autoregressive parameters. For the same model, we compare the finite sample properties of various test statistics for linear restrictions on autoregressive parameters. These tests include the standard asymptotic Wald test based on various GMMEs, a bootstrapped version of the Wald test, two versions of the C(α) test, the standard Lagrange multiplier (LM) test, the minimum chi-square test (MC), and two versions of the generalized method of moments (GMM) criterion test. Finally, we study the finite sample properties of effects estimators that show how changes in explanatory variables impact the dependent variable.     

Disclosure and Relationship Satisfaction in Families

The present study developed a theoretical framework for understanding the social mechanisms underlying disclosure and its link with relationship satisfaction in a full family design. A study among 262 intact families, consisting of 2 children and 2 adults and involving 1,048 individuals, applied the social relations model. Results showed that disclosure was more important to satisfaction in horizontal relationships than vertical ones. Further, relationship‐specific disclosure was more important to satisfaction than dispositional disclosure. These results have implications for the examination of relationship regulation and maintenance in (non)voluntary relationships and the development of psychosocial problems in parent‐ child relationships.     

Balancing work and welfare: activation and flexicurity policies in The Netherlands, 1980–2000

As a result of the flexibilisation of labour and the trend towards the 'activating welfare state', social policies show an increasing interconnection of work and welfare issues. The Netherlands is no exception. It is generally believed that the Dutch welfare state is successfully activating its unemployed labour potential (often referred to as the 'Dutch Miracle'), and that flexible and part-time work is protected by adequate 'flexicurity'. This article critically reviews Dutch activation and flexicurity policies. It concludes that there is still more unemployment than the miracle-story suggests; that important target groups of activation policies have not profited from 'the miracle'; that part-time workers have sufficient social protection but that social security for flex-workers still needs major improvements, despite favourable adjustments of labour law.     

Social Influence,Risk and Benefit Perceptions,and the Acceptability of Risky Energy Technologies: An Explanatory Model of Nuclear Power Versus Shale Gas

Risky energy technologies are often controversial and debates around them are polarized; in such debates public acceptability is key. Research on public acceptability has emphasized the importance of intrapersonal factors but has largely neglected the influence of interpersonal factors. In an online survey (N = 948) with a representative sample of the United Kingdom, we therefore integrate interpersonal factors (i.e., social influence as measured by social networks) with two risky energy technologies that differ in familiarity (nuclear power vs. shale gas) to examine how these factors explain risk and benefit perceptions and public acceptability. Findings show that benefit perceptions are key in explaining acceptability judgments. However, risk perceptions are more important when people are less familiar with the energy technology. Social network factors affect perceived risks and benefits associated with risky energy technology, hereby indirectly helping to form one's acceptability judgment toward the technology. This effect seems to be present regardless of the perceived familiarity with the energy technology. By integrating interpersonal with intrapersonal factors in an explanatory model, we show how the current “risk–benefit acceptability” model used in risk research can be further developed to advance the current understanding of acceptability formation.     

Learning causal structure from mixed data with missing values using Gaussian copula models

We consider the problem of causal structure learning from data with missing values, assumed to be drawn from a Gaussian copula model. First, we extend the ‘Rank PC’ algorithm, designed for Gaussian copula models with purely continuous data (so-called nonparanormal models), to incomplete data by applying rank correlation to pairwise complete observations and replacing the sample size with an effective sample size in the conditional independence tests to account for the information loss from missing values. When the data are missing completely at random (MCAR), we provide an error bound on the accuracy of ‘Rank PC’ and show its high-dimensional consistency. However, when the data are missing at random (MAR), ‘Rank PC’ fails dramatically. Therefore, we propose a Gibbs sampling procedure to draw correlation matrix samples from mixed data that still works correctly under MAR. These samples are translated into an average correlation matrix and an effective sample size, resulting in the ‘Copula PC’ algorithm for incomplete data. Simulation study shows that: (1) ‘Copula PC’ estimates a more accurate correlation matrix and causal structure than ‘Rank PC’ under MCAR and, even more so, under MAR and (2) the usage of the effective sample size significantly improves the performance of ‘Rank PC’ and ‘Copula PC.’ We illustrate our methods on two real-world datasets: riboflavin production data and chronic fatigue syndrome data.

     

Assessing the predictive value of a binary surrogate for a binary true endpoint based on the minimum probability of a prediction error

The individual causal association (ICA) has recently been introduced as a metric of surrogacy in a causal‐inference framework. The ICA is defined on the unit interval and quantifies the association between the individual causal effect on the surrogate (ΔS) and true (ΔT) endpoint. In addition, the ICA offers a general assessment of the surrogate predictive value, taking value 1 when there is a deterministic relationship between ΔT and ΔS, and value 0 when both causal effects are independent. However, when one moves away from the previous two extreme scenarios, the interpretation of the ICA becomes challenging. In the present work, a new metric of surrogacy, the minimum probability of a prediction error (PPE), is introduced when both endpoints are binary, ie, the probability of erroneously predicting the value of ΔT using ΔS. Although the PPE has a more straightforward interpretation than the ICA, its magnitude is bounded above by a quantity that depends on the true endpoint. For this reason, the reduction in prediction error (RPE) attributed to the surrogate is defined. The RPE always lies in the unit interval, taking value 1 if prediction is perfect and 0 if ΔS conveys no information on ΔT. The methodology is illustrated using data from two clinical trials and a user‐friendly R package Surrogate is provided to carry out the validation exercise.     

Non‐take up of the supplemental child benefit for children with a disability in Belgium: A mixed‐method approach

Families with disabled children run a great risk of being poor. Although policies providing poor families with financial benefits should be effective in reducing poverty, the actual effectiveness is often jeopardized by the issue of non‐take up (NTU). Yet, NTU of benefits aimed at disabled children is for the most part uncharted territory. In this article, we fill this gap using a mixed‐methods approach to (a) estimate the magnitude and characteristics of NTU in the Belgian “supplemental child benefit” by drawing on a large‐scale administrative dataset on childhood disabilities and (b) explore the determinants of NTU by means of semistructured interviews with experts and parents. We estimate an NTU rate of at least 10%, a substantial figure given that the benefit is not income tested. This mainly concerns children with “less visible disabilities” (autism spectrum disorder and other intellectual and psychological disorders) and results from insufficient information provision about the benefit's existence and eligibility criteria; process costs, for instance, the complexity of the procedure; and the way the scale to assess a child's disability is constructed.     

Conversion of Verbal Response Scales: Robustness Across Demographic Categories

Happiness and life satisfaction have traditionally been measured using verbal response scales, however, these verbal scales have not kept up with the present trend to use numerical response scales. A switch from a verbal scale to a numerical scale, however, causes a severe problem for trend analyses, due to the incomparability of the old and new measurements. The Reference Distribution Method is a method that has been developed recently to deal with this comparison problem. In this method use is made of a reference distribution based on responses to a numerical scale which is used to decide at which point verbally labelled response options transit from one state to another, for example from ‘happy’ to ‘very happy’. Next, for each wave of the time series in which the verbal scale is used, a population mean is estimated for the beta distribution that fits best to these transition points and the responses in this wave. These estimates are on a level that is comparable to that of the mean of the reference distribution and are appropriate for use in an extended time series based on the responses measured using a verbal and a numerical scale. In this paper we address the question of whether the transition points derived for the general population can be used for demographic categories to produce reliable, extended time series to monitor differences in trends among these categories. We conclude that this is possible and that it is not necessary to derive transition points for each demographic category separately.