Options
Vallée, Audrey-Anne
Nom
Vallée, Audrey-Anne
Affiliation principale
Fonction
Ancien.ne collaborateur.trice
Identifiants
Résultat de la recherche
Voici les éléments 1 - 10 sur 11
- PublicationAccès libreLinearisation for Variance Estimation by Means of Sampling Indicators: Application to Non‐response(2019-8-19)
; In order to estimate the variance of estimators in survey sampling, we consider a method in which the estimators are linearized with respect to the basic random variables: the sampling indicator and the response indicator. When a superpopulation model is assumed, the estimators can also be linearized with respect to the variable of interest. This method ensures the derivation of a variance since the estimated parameters are linearized with respect to the random variables directly. It becomes particularly straightforward to construct explicit variance estimators. All sources of randomness are taken into account. The effects caused by the complexity of the estimation method, the calibration and the nonresponse treatment, imputation or reweighting, appear automatically and explicitly in the linearization variables. Through a set of examples, we show the simplicity of the method. Some results regarding the estimation of variance with nonresponse can be obtained in a simpler way than the usual developments. A set of simulations illustrates the proposed methodology. - PublicationMétadonnées seulement
- PublicationMétadonnées seulementBalanced imputation for swiss cheese nonresponse(2018-9-20)
; Swiss cheese nonresponse or non-monotone nonresponse occurs when all the variables of a survey can contain missing values without a particular pattern. Imputation of missing values allows to reduce the bias and the variability due to nonresponse. It is difficult to preserve the distributions and the relations between the variables when imputing in the swiss cheese nonresponse case. In this presentation, balanced K-nearest neighbor imputation Hasler and Tillé (2016) is extended to treat swiss cheese nonresponse. It is a donor imputation method that is random and constructed to meet some requirements. First, a nonrespondent can be imputed by donors which are close to him. The distances are calculated with the observed values. Next, all the missing values of a nonrespondent are imputed by the same donor. Last, the donors are chosen so that if the observed values of the nonrespondents were imputed, the estimated totals would be the same as the one calculated with the observed values only. To meet all the requirements, a matrix of imputation probabilities is constructed with calibration techniques. The donors are selected with these imputation probabilities and balanced sampling methods. The advantages and the properties of the method are investigated in a simulation study. - PublicationMétadonnées seulement
- PublicationMétadonnées seulement
- PublicationAccès libreRevisiting Variance Decomposition when Independent Samples Intersect(2017-7-21)
; The variance and the estimated variance of the expanded estimator in the intersection of two independent samples can be decomposed into two ways. Due to the inclusion probabilities, it is generally more practical to compute the variance with one decomposition. With the other one, it is more convenient to estimate the variance. - PublicationMétadonnées seulementBalanced Imputation for Swiss Cheese Nonresponse(2017-6-10)
; Swiss cheese nonresponse refers to the case where all the variables can contain missing values in a general pattern even if this wording is abuse since most of the Swiss cheeses do not have holes. However, in the case of Swiss cheese nonresponse, it is not possible to consider that some variables are known for every units and can thus be used as auxiliary variables. We propose a new technique of random donor imputation. The method is based on the establishment of consistency principles. A nonrespondent and its donor should be close to each other. The same donor must be used for all missing variables of a unit. Moreover we impose that, if we were imputing the known variables of a nonrespondent by the values of the donor, the totals of these variables must remain the same. The procedure is based on the calculation of imputation probabilities. Next, the donors are selected randomly in such a way that the constraints are satisfied. - PublicationMétadonnées seulementEstimation de la variance par linéarisation via l'indicatrice d'échantillonnage avec application à la non-réponse(2016-10-13)
; En présence de non-réponse, la repondération et l'imputation sont couramment utilisées pour l'estimation de paramètres d'intérêts. Un éventail de méthodes est détaillé dans la littérature. À l'étape de l'inférence, plusieurs aspects sont à considérer: le mécanisme de non-réponse, le cadre de travail pour l'inférence (basé sur le modèle de non-réponse ou sur le modèle d'imputation) et la méthode d'imputation. Ces aspects sont importants pour l'estimation de la variance du paramètre d'intérêt. Cette variance est souvent approchée à l'aide d'une linéarisation de Taylor par rapport aux totaux ou aux poids de sondage. Dans cette présentation, une approche est utilisée pour linéariser directement l'estimateur du paramètre d'intérêt par rapport aux éléments aléatoires. Cette approche permet de simplifier les calculs et d'obtenir un estimateur de la variance explicite. Cette technique est appliquée aux deux cadres de travail et à différentes méthodes d'imputation. Elle permet aussi l'estimation de la variance de statistiques non-linéaires dans une base de données complète, par exemple celle de l'estimateur calé. - PublicationMétadonnées seulementIncorporating Spatial and Operational Constraints in the Sampling Designs for Forest Inventories(2015-9-3)
; - PublicationMétadonnées seulementIncorporating spatial and operational constraints in the sampling designs for forest inventories(2015-7-15)
;