Suchergebnisse

"This paper discusses the problem of creating general purpose calibrated survey weights when the control totals data exist at different levels of aggregation, such as households and individuals. We present and compare three different methods. The first does the weighting in two stages, using only the household data, and then only the individual data. The second redefines targets at the individual level, if possible, and uses these targets to calibrate only the individual level weights. The third uses multipliers of household size to produce household level weights that simultaneously calibrate to the individual level totals. We discuss the advantages and disadvantages of these approaches, including control total data accessibility and available software from the perspective a survey statistician working outside of a national statistical organization. We conclude by outlining directions for further research." (author's abstract)

Survey response rates for modern surveys using many different modes are trending downward leaving the potential for nonresponse biases
in estimates derived from using only the respondents. The reasons for nonresponse may be complex functions of known auxiliary variables or
unknown latent variables not measured by practitioners. The degree to which the propensity to respond is associated with survey outcomes
casts light on the overall potential for nonresponse biases for estimates of means and totals. The most common method for nonresponse
adjustments to compensate for the potential bias in estimates has been logistic and probit regression models. However, for more complex
nonresponse mechanisms that may be nonlinear or involve many interaction effects, these methods may fail to converge and thus fail to
generate nonresponse adjustments for the sampling weights. In this paper we compare these traditional techniques to a relatively new data
mining technique- random forests – under a simple and complex nonresponse propensity population model using both direct and propensity
stratification nonresponse adjustments. Random forests appear to offer marginal improvements for the complex response model over logistic
regression in direct propensity adjustment, but have some surprising results for propensity stratification across both response models.

Simulating Distributional Impacts of Macro-dynamics: Theory and Practical Applications is a comprehensive guide for analyzing and understanding the effects of macroeconomic shocks on income and consumption distribution, as well as using the ADePT Simulation Module. Since real-time micro data is rarely available, the Simulation Module (part of the ADePT economic analysis software) takes advantage of historical household surveys to estimate how current or proposed macro changes might impact household and individuals welfare. Using examples from different economic and social contexts, the book ex

Cet article propose une vision du marché du travail russe permettant d'expliquer la coexistence d'éléments d'inertie et de flexibilité par un phénomène de segmentation. Plaçant l'incertitude au cœur des décisions des firmes et des employés, il montre que les travailleurs, qui ont tous une aversion pour le risque, peuvent adopter des comportements différenciés selon leur productivité relative. Certains, bénéficiant de perspectives favorables, adoptent un comportement mobile et nouent des contrats prévoyant de les rémunérer en fonction de leur productivité. D'autres, jouissant d'opportunités moins favorables, acceptent une réduction de leur rémunération réelle en échange de la sécurité de l'emploi et de l'accès aux actifs sociaux de la firme. Notre cadre d'analyse permet de proposer une interprétation des arriérés de salaire. Ces derniers constitueraient un élément du contrat d'assurance implicite noué entre l'entreprise et certains travailleurs. Nous vérifions les prévisions du modèle à l'aide d'une base de données comprenant 13 628 entreprises russes, observées de 1992 à 1997.

Held in October 2018, The Big Data Meets Survey Science conference, also known as "BigSurv18," provided a first-of-its-kind opportunity for survey researchers, statisticians, computer scientists, and data scientists to convene under the same roof. At this conference, scientists from multiple disciplines were able to exchange ideas about their work might influence and enhance the work of others. This was a landmark event, especially for survey researchers and statisticians, whose industry has been buffeted of late by falling response rates and rising costs at the same time as a proliferation of new tools and techniques, coupled with increasing availability of data, has resulted in "Big Data" approaches to describing and modelling human behavior.