Results 1  10
of
10
Multiple imputation for multivariate missingdata problems: a data analyst's perspective
 Multivariate Behavioral Research
, 1998
"... Analyses of multivariate data are frequently hampered by missing values. Until recently, the only missingdata methods available to most data analysts have been relatively ad hoc practices such as listwise deletion. Recent dramatic advances in theoretical and computational statistics, however, hav ..."
Abstract

Cited by 49 (1 self)
 Add to MetaCart
Analyses of multivariate data are frequently hampered by missing values. Until recently, the only missingdata methods available to most data analysts have been relatively ad hoc practices such as listwise deletion. Recent dramatic advances in theoretical and computational statistics, however, have produced a new generation of flexible procedures with a sound statistical basis. These procedures involve multiple imputation (Rubin, 1987), a simulation technique that replaces each missing datum with a set of m>1 plausible values. The m versions of the complete data are analyzed by standard completedata methods, and the results are combined using simple rules to yield estimates, standard errors, and pvalues that formally incorporate missingdata uncertainty. New computational algorithms and software described in a recent book (Schafer, 1997) allow us to create proper multiple imputations in complex multivariate settings. This article reviews the key ideas of multiple imputation, discusses the software programs currently available, and demonstrates their use on data from
Diagnostics for multivariate imputations
 Applied Statistics
, 2009
"... Summary. We consider three sorts of diagnostics for random imputations: displays of the completed data, which are intended to reveal unusual patterns that might suggest problems with the imputations, comparisons of the distributions of observed and imputed data values and checks of the fit of observ ..."
Abstract

Cited by 7 (2 self)
 Add to MetaCart
Summary. We consider three sorts of diagnostics for random imputations: displays of the completed data, which are intended to reveal unusual patterns that might suggest problems with the imputations, comparisons of the distributions of observed and imputed data values and checks of the fit of observed data to the model that is used to create the imputations. We formulate these methods in terms of sequential regression multivariate imputation, which is an iterative procedure in which the missing values of each variable are randomly imputed conditionally on all the other variables in the completed data matrix. We also consider a recalibration procedure for sequential regression imputations. We apply these methods to the 2002 environmental sustainability index, which is a linear aggregation of 64 environmental variables on 142 countries.
How to Deal with Missing Categorical Data: Test of a Simple Bayesian Method
 Organizational Research Methods
, 2003
"... ..."
Balanced Repeated Replication For Stratified Multistage Survey Data Under Imputation
"... Balanced repeated replication (BRR) is a popular method for variance estimation in surveys. The standard BRR method works by first creating a set of "balanced" pseudoreplicated data sets from the original data set. For a survey estimator `, the BRR variance estimator is the average of sq ..."
Abstract

Cited by 2 (0 self)
 Add to MetaCart
Balanced repeated replication (BRR) is a popular method for variance estimation in surveys. The standard BRR method works by first creating a set of "balanced" pseudoreplicated data sets from the original data set. For a survey estimator `, the BRR variance estimator is the average of squared deviations ` (r) \Gamma `, where ` (r) is the same as ` but based on the data in the rth pseudoreplicated data set only. When there are a large number of imputed missing values (nonrespondents), however, treating the imputed values as observed data and applying the standard BRR variance estimation formula does not produce valid variance estimators. Intuitively, the variation due to imputation can be captured by the BRR method if every pseudoreplicated data set is imputed in exactly the same way as the original data set is imputed (assuming that the data set contains flags for nonrespondents). When a random imputation method (such as random hot deck imputation, random ratio imputation,...
Inference With Survey Data Imputed By Hot Deck When Nonrespondents Are Nonidentifiable
"... Hot deck imputation for nonrespondents is often used in surveys. It is a common practice to treat the imputed values as if they are true values, and compute survey estimators and their variance estimators using standard formulas. The variance estimators, however, have seriously negative biases when ..."
Abstract
 Add to MetaCart
Hot deck imputation for nonrespondents is often used in surveys. It is a common practice to treat the imputed values as if they are true values, and compute survey estimators and their variance estimators using standard formulas. The variance estimators, however, have seriously negative biases when the rate of nonresponse is appreciable. Methods such as the multiple imputation and the adjusted jackknife have been proposed to obtain improved variance estimators. However, multiple imputation requires that multiple data sets be generated and maintained and that the imputation procedure be proper; the adjusted jackknife requires "flags" to identify nonrespondents. In many practical problems there is only a single imputed data set with unknown response status (no identification flag). In this paper we derive some asymptotically designconsistent inference procedures in the situation where a stratified multistage sampling design is used to collect survey data; hot deck imputation is applied ...
USE OF ADMINISTRATIVE DATA TO EXPLORE EFFECT OF ESTABLISHMENT NONRESPONSE ADJUSTMENT ON THE NATIONAL COMPENSATION SURVEY ESTIMATES December 2006
"... Unit nonresponse is a well known but undesirable problem in sample surveys including the National Compensation Survey (NCS) Program. In NCS, unit ..."
Abstract
 Add to MetaCart
Unit nonresponse is a well known but undesirable problem in sample surveys including the National Compensation Survey (NCS) Program. In NCS, unit
Direct Prediction Methods on Lifetime Distribution of Organic LightEmitting Diodes From Accelerated Degradation Tests
"... Abstract—Accelerated degradation testing (ADT) expedites product degradation by stressing the product beyond its normal use. To extrapolate the product’s reliability at use condition, the ADT requires a known functional link relating the harsh testing environment to the usual use environment. Practi ..."
Abstract
 Add to MetaCart
Abstract—Accelerated degradation testing (ADT) expedites product degradation by stressing the product beyond its normal use. To extrapolate the product’s reliability at use condition, the ADT requires a known functional link relating the harsh testing environment to the usual use environment. Practitioners are often faced with a great challenge to designate an explicit form of the stressdegradation relationship a priori in accelerated degradation models. In this paper, we propose three methods to make direct inference on the lifetime distribution itself without invoking arbitrary assumptions on the degradation model: delta approximation, multiple imputation of failuretimes, and the lifetime distributionbased (LDB) method. The methods are easy to implement without computational difficulty, hence they have potential in a wide range of applications for estimating lifetime distributions from ADT data. We applied the methods to two ADT data sets including a real application of commercial organic lightemitting diodes (OLED). The analysis of the examples and simulation results suggests parametric LDB and multiple imputation method as more potential alternatives to traditional failuretime approaches, especially for the case where there is neither enough physical background, nor historical evidence supporting presumed relationships between stress and the parameters of the degradation model. Index Terms—Accelerated degradation test, delta approximation, lifetime distributionbased procedure, multiple imputation, nonlinear randomcoefficients model, organic lightemitting diode. ACRONYM1 ADT accelerated degradation testing ALT accelerated life testing
Update on Use of Administrative Data to Explore Effect of Establishment Nonresponse Adjustment on the National Co mpensation Survey Estimates October 2008
"... Nearly all establishment surveys are prone to some level of nonresponse. Nonresponse may lead to biases in survey estimates and an increase in survey sampling variance. Survey practitioners use various techniques to reduce bias due to nonresponse. The most common technique is to adjust the sampli ..."
Abstract
 Add to MetaCart
Nearly all establishment surveys are prone to some level of nonresponse. Nonresponse may lead to biases in survey estimates and an increase in survey sampling variance. Survey practitioners use various techniques to reduce bias due to nonresponse. The most common technique is to adjust the sampling weights of responding units to account for nonresponding units within a specified set of weighting classes or cells. In the National Compensation Survey (NCS), which is an establishment survey conducted by the Bureau of Labor Statistics, the weighting cells are formed using available auxiliary information: ownership, industry, and establishment employment size. At JSM 2006, we presented a paper in which we explored how effective the formed cells are in reducing potential bias in the NCS estimates and presented results for one NCS area. Since 2006, NCS has expanded the study of potential bias due to nonresponse to several additional survey areas and time periods. In this paper, we present results from this additional research. We include localities of different size and with different levels of nonresponse. Also we compare the direction and magnitude of bias across time and across areas.
MULTIPLE IMPUTATION OF INDUSTRY AND OCCUPATION CODES FOR PUBLICUSE FILES 1
"... I. INTRODUCTION imputed rather than true values is ignored, For each decennial census in the United States, the responses concerning employment are classif ied into industry and occupation categories. The industry and occupation coding ..."
Abstract
 Add to MetaCart
I. INTRODUCTION imputed rather than true values is ignored, For each decennial census in the United States, the responses concerning employment are classif ied into industry and occupation categories. The industry and occupation coding