Question Order Effects on Subjective Measures of Quality of Life1

The effects of question order on respondents' ratings of general and specific aspects of community life were assessed using data from separate mail surveys in Montana and Pennsylvania, The samples differed in locale, composition, and size and the relevant questions varied in number, format, and specific focus. Nevertheless, for both data sets the general question was more likely to be answered and more likely to receive positive responses when it was asked after—rather than before—the specific questions. There was some indication that carryover from the specific items to the general question responses were somewhat greater for those questions asked most recently and less for those asked earlier. Implications of these findings are discussed.     

Maximum Likelihood Inference for Log-linear Models Subject to Constraints of Double Monotone Dependence

To model an hypothesis of double monotone dependence between two ordinal categorical variables A and B usually a set of symmetric odds ratios defined on the joint probability function is subject to linear inequality constraints. Conversely in this paper two sets of asymmetric odds ratios defined, respectively, on the conditional distributions of A given B and on the conditional distributions of B given A are subject to linear inequality constraints. If the joint probabilities are parameterized by a saturated log-linear model, these constraints are nonlinear inequality constraints on the log-linear parameters. The problem here considered is a non-standard one both for the presence of nonlinear inequality constraints and for the fact that the number of these constraints is greater than the number of the parameters of the saturated log-linear model.This work has been supported by the COFIN 2002 project, references 2002133957_002, 2002133957_004. Preliminary findings have been presented at SIS (Società Italiana di Statistica) Annual Meeting, Bari, 2004.     

Age difference asymmetry and a two-sex perspective

The paper shows that (a) the age difference between spouses can be and has been different for men and women; (b) male and female age differences can exhibit different trends; (c) the discrepancy is due to remarriage; (d) the age difference fluctuates substantially through time; (e) trends in the age difference are not readily interpretable as reflecting change in the relative status of the sexes; and (f) that trends in the age difference may be linked to marriage market conditions. The paper argues that marriage and fertility data for both men and women are increasingly necessary because of rises in marital breakdown and the associated divergence of the experience of men and women.Cet article montre que (a) les différences d'âge entre époux peuvent être et ont été dissemblables pour les hommes et les femmes; (b) les différences d'âge masculines et féminines peuvent présenter des tendances différentes; (c) la divergence est due aux remariages; (d) les différences d'âge varient de façon importante au cours du temps; (e) les tendances dans les différences d'âge ne sont pas simplement interprétables en termes de changements dans le statut relatif des sexes; et (f) les tendances dans les différences d'âge peuvent être reliées aux conditions du marché matrimonial. L'article montre que les données sur le mariage et la fécondité à la fois des hommes et des femmes, sont de plus en plus nécessaires du fait de l'accroissement des ruptures d'unions et de la divergence des expériences masculines et féminines, qui lui est associée.

---

An earlier version of this paper, His and her age gap: asymmetry in the age difference between partners, was a contributed paper to Session F27 of the General Conference of the International Union for the Scientific Study of Population, New Delhi, 20–27 September 1989. A longer version appeared as LS Working Paper 70, issued by City University Social Statistics Research Unit, September 1990.     

On Combining Wavelets Expansion and Sparse Linear Models for Regression on Metabolomic Data and Biomarker Selection

Wavelet thresholding of spectra has to be handled with care when the spectra are the predictors of a regression problem. Indeed, a blind thresholding of the signal followed by a regression method often leads to deteriorated predictions. The scope of this article is to show that sparse regression methods, applied in the wavelet domain, perform an automatic thresholding: the most relevant wavelet coefficients are selected to optimize the prediction of a given target of interest. This approach can be seen as a joint thresholding designed for a predictive purpose. The method is illustrated on a real world problem where metabolomic data are linked to poison ingestion. This example proves the usefulness of wavelet expansion and the good behavior of sparse and regularized methods. A comparison study is performed between the two-steps approach (wavelet thresholding and regression) and the one-step approach (selection of wavelet coefficients with a sparse regression). The comparison includes two types of wavelet bases, various thresholding methods, and various regression methods and is evaluated by calculating prediction performances. Information about the location of the most important features on the spectra was also obtained and used to identify the most relevant metabolites involved in the mice poisoning.     

A Note on Selection and Gender Unemployment Gaps

Previous research has documented the impact of positive selection into employment on the ranking of countries by gender wage gaps. This note focuses on the impact of selection into labor force on cross-country differences in gender unemployment gaps. We construct the Manski bounds for the selection-free gender unemployment gaps in 26 EU countries and show that - without additional assumptions - the observed gender unemployment gaps carry little information about the selection-free gender differences in unemployment. Contrary to the common assumption of positive selection into labor force (similar to positive selection into employment documented in the gender wage gap research), we also point at an example of negative-selection bias. We show that labor force withdrawal of mothers on job-protected family leaves may lead to an overestimation of the selection-free gender unemployment gaps by as much as 1 p.p.     

Multilevel maximum likelihood estimation with application to covariance matrices

The asymptotic variance of the maximum likelihood estimate is proved to decrease when the maximization is restricted to a subspace that contains the true parameter value. Maximum likelihood estimation allows a systematic fitting of covariance models to the sample, which is important in data assimilation. The hierarchical maximum likelihood approach is applied to the spectral diagonal covariance model with different parameterizations of eigenvalue decay, and to the sparse inverse covariance model with specified parameter values on different sets of nonzero entries. It is shown computationally that using smaller sets of parameters can decrease the sampling noise in high dimension substantially.