Bayesian parameter estimation via variational methods

We consider a logistic regression model with a Gaussian prior distribution over the parameters. We show that an accurate variational transformation can be used to obtain a closed form approximation to the posterior distribution of the parameters thereby yielding an approximate posterior predictive model. This approach is readily extended to binary graphical model with complete observations. For graphical models with incomplete observations we utilize an additional variational transformation and again obtain a closed form approximation to the posterior. Finally, we show that the dual of the regression problem gives a latent variable density model, the variational formulation of which leads to exactly solvable EM updates.     

A TEST OF COMPETING EXPLANATIONS OF COMPENSATION DEMANDED

We find that prospect theory behavior is significantly more prevalent than utility theory behavior in experiments involving multiple, real items. In the experiments, subjects were endowed with three items and asked the minimum payments they required to be willing to return one, two, or three of them. Our key observation is that prospect theory implies concavity of compensation demanded, whereas utility theory implies convexity. We examine whether the compensation demanded is convex or concave in the number of items returned. ( JEL C91)     

Social capital and child welfare

Social and economic inequalities have risen steeply in the UK over the past 20 years. Children and their families have been particularly disadvantaged by this creeping impoverishment, which is associated with negative effects on parenting capacity and developmental outcomes for children. The social capital of communities, which consists of the cultural resources and inter‐personal relationships between members, is also eroded by inequality and social exclusion. Evidence is presented that demonstrates the way in which children's welfare and family functioning are crucially dependent upon the social support available within local communities. It is argued that building social capital in poor communities is a more effective way of promoting children's welfare than the present emphasis on formal child protection and family support services and efforts to increase parenting skills and responsibilities. Copyright © 1999 John Wiley & Sons, Ltd.     

Testing a Parametric Model Against a Nonparametric Alternative with Identification Through Instrumental Variables

This paper is concerned with inference about a function g that is identified by a conditional moment restriction involving instrumental variables. The paper presents a test of the hypothesis that g belongs to a finite‐dimensional parametric family against a nonparametric alternative. The test does not require nonparametric estimation of g and is not subject to the ill‐posed inverse problem of nonparametric instrumental variables estimation. Under mild conditions, the test is consistent against any alternative model. In large samples, its power is arbitrarily close to 1 uniformly over a class of alternatives whose distance from the null hypothesis is O(n^−1/2), where n is the sample size. In Monte Carlo simulations, the finite‐sample power of the new test exceeds that of existing tests.     

Protective indigenous collective value of Ubuntu and child neglect: Implications for rural child protection practice

Theories on collective efficacy and social support suggest that indigenous values that support collective practices and sanction community obligations to childcare would be protective against child neglect. Likewise, new qualitative findings show that collective values are stronger in rural areas than in urban. This study tested the claims that the value of Ubuntu, which is a symbolic cultural value of ‘being for others’, will be protective against the likelihood of neglect; this relationship will be stronger in rural compared with urban communities in Ghana. Using data obtained from a nationally representative sample of 1100 mothers (from 22 communities) in Ghana, we tested the claims using fixed effects logistic regression. The Ubuntu norms were significantly endorsed in rural communities compared with the urban. The overall model showed that higher levels of Ubuntu are associated with lower odds of child neglect (OR .47, [.29, .76] p < 0.05), and the relationship remained significant only in the rural sample (OR .13, [.06, .31] p < 0.001). Similar evidence was recorded for the Ubuntu norms of community care and compassion. The results suggest that child protection in rural Ghana can be fruitful when interventions are developed to boost the value of Ubuntu and the norms of collective childcare.     

Does IO Economics Really Make the Case for Investments in IT: Comment on West and Courtney

West and Courtney [18] rely on industrial organization economics in general, and the economics of uncertainty in particular, for insights into the benefits of overcoming information constraints through information systems (IS) and improved information technologies (IT). This Comment argues that the IS researcher who looks to the economics of uncertainty to make a case for improved IT will only find support from a selective reading of a constantly expanding literature whose aims and inferences are easily misunderstood.     

‘You've got to trust her and she's got to trust you’: children's views on participation in the child protection system

Participation by children in child protection remains a complex area of practice. This paper presents findings from a qualitative study exploring the views of 26 children, aged 6–17 years, about their participation in the child protection system in England. All of the children were subject to a child protection plan and were living at home at the time of interview. The children's understanding of the child protection process was categorized, and the majority of children, including the youngest, were found be at least partially aware of the child protection process, often struggling to make sense of the professional intervention in their families on the basis of partial information. It is argued that decisions about children's involvement should take into account not only children's age and understanding, but be seen in the context of wider family dynamics. Participation in formal processes such as child protection conferences was experienced as difficult and emotive. The child's relationship with their social worker was central to meaningful participation.     

VARIABLE SELECTION IN NONPARAMETRIC ADDITIVE MODELS

We consider a nonparametric additive model of a conditional mean function in which the number of variables and additive components may be larger than the sample size but the number of nonzero additive components is "small" relative to the sample size. The statistical problem is to determine which additive components are nonzero. The additive components are approximated by truncated series expansions with B-spline bases. With this approximation, the problem of component selection becomes that of selecting the groups of coefficients in the expansion. We apply the adaptive group Lasso to select nonzero components, using the group Lasso to obtain an initial estimator and reduce the dimension of the problem. We give conditions under which the group Lasso selects a model whose number of components is comparable with the underlying model, and the adaptive group Lasso selects the nonzero components correctly with probability approaching one as the sample size increases and achieves the optimal rate of convergence. The results of Monte Carlo experiments show that the adaptive group Lasso procedure works well with samples of moderate size. A data example is used to illustrate the application of the proposed method.     

Joint covariate selection and joint subspace selection for multiple classification problems

We address the problem of recovering a common set of covariates that are relevant simultaneously to several classification problems. By penalizing the sum of ℓ ₂ norms of the blocks of coefficients associated with each covariate across different classification problems, similar sparsity patterns in all models are encouraged. To take computational advantage of the sparsity of solutions at high regularization levels, we propose a blockwise path-following scheme that approximately traces the regularization path. As the regularization coefficient decreases, the algorithm maintains and updates concurrently a growing set of covariates that are simultaneously active for all problems. We also show how to use random projections to extend this approach to the problem of joint subspace selection, where multiple predictors are found in a common low-dimensional subspace. We present theoretical results showing that this random projection approach converges to the solution yielded by trace-norm regularization. Finally, we present a variety of experimental results exploring joint covariate selection and joint subspace selection, comparing the path-following approach to competing algorithms in terms of prediction accuracy and running time.