Survey Methodology

View the most recent version.

June 2011

Content note: At this moment, full content is available in PDF only. A print version can also be ordered.

To access the PDF publication, please use the "Full content in PDF" link on the sidebar (on the left-hand side of this page).

Accessibility notice: If the following information is not accessible to you, please contact us for assistance or to obtain it in alternate formats.

---

The journal Survey Methodology Volume 37, Number 1 (June 2011) contains the following 9 papers:

Regular Papers:

Nonsampling errors in dual frame telephone surveys

J. Michael Brick, Ismael Flores Cervantes, Sunghee Lee and Greg Norman

Abstract

Dual frame telephone surveys are becoming common in the U.S. because of the incompleteness of the landline frame as people transition to cell phones. This article examines nonsampling errors in dual frame telephone surveys. Even though nonsampling errors are ignored in much of the dual frame literature, we find that under some conditions substantial biases may arise in dual frame telephone surveys due to these errors. We specifically explore biases due to nonresponse and measurement error in these telephone surveys. To reduce the bias resulting from these errors, we propose dual frame sampling and weighting methods. The compositing factor for combining the estimates from the two frames is shown to play an important role in reducing nonresponse bias.

Maximum likelihood estimation for contingency tables and logistic regression with incorrectly linked data

James O. Chipperfield, Glenys R. Bishop and Paul Campbell

Abstract

Data linkage is the act of bringing together records that are believed to belong to the same unit (e.g., person or business) from two or more files. It is a very common way to enhance dimensions such as time and breadth or depth of detail. Data linkage is often not an error-free process and can lead to linking a pair of records that do not belong to the same unit. There is an explosion of record linkage applications, yet there has been little work on assuring the quality of analyses using such linked files. Naively treating such a linked file as if it were linked without errors will, in general, lead to biased estimates. This paper develops a maximum likelihood estimator for contingency tables and logistic regression with incorrectly linked records. The estimation technique is simple and is implemented using the well-known EM algorithm. A well known method of linking records in the present context is probabilistic data linking. The paper demonstrates the effectiveness of the proposed estimators in an empirical study which uses probabilistic data linkage.

Hierarchical Bayes small area estimation under a spatial model with application to health survey data

Yong You and Qian M. Zhou

Abstract

In this paper we study small area estimation using area level models. We first consider the Fay-Herriot model (Fay and Herriot 1979) for the case of smoothed known sampling variances and the You-Chapman model (You and Chapman 2006) for the case of sampling variance modeling. Then we consider hierarchical Bayes (HB) spatial models that extend the Fay-Herriot and You-Chapman models by capturing both the geographically unstructured heterogeneity and spatial correlation effects among areas for local smoothing. The proposed models are implemented using the Gibbs sampling method for fully Bayesian inference. We apply the proposed models to the analysis of health survey data and make comparisons among the HB model-based estimates and direct design-based estimates. Our results have shown that the HB model-based estimates perform much better than the direct estimates. In addition, the proposed area level spatial models achieve smaller CVs than the Fay-Herriot and You-Chapman models, particularly for the areas with three or more neighbouring areas. Bayesian model comparison and model fit analysis are also presented.

Small area estimation under transformation to linearity

Hukum Chandra and Ray Chambers

Abstract

Small area estimation based on linear mixed models can be inefficient when the underlying relationships are non-linear. In this paper we introduce SAE techniques for variables that can be modelled linearly following a non-linear transformation. In particular, we extend the model-based direct estimator of Chandra and Chambers (2005, 2009) to data that are consistent with a linear mixed model in the logarithmic scale, using model calibration to define appropriate weights for use in this estimator. Our results show that the resulting transformation-based estimator is both efficient and robust with respect to the distribution of the random effects in the model. An application to business survey data demonstrates the satisfactory performance of the method.

The construction of stratified designs in R with the package stratification

Sophie Baillargeon and Louis-Paul Rivest

Abstract

This paper introduces a R-package for the stratification of a survey population using a univariate stratification variable X and for the calculation of stratum sample sizes. Non iterative methods such as the cumulative root frequency method and the geometric stratum boundaries are implemented. Optimal designs, with stratum boundaries that minimize either the CV of the simple expansion estimator for a fixed sample size n or the n value for a fixed CV can be constructed. Two iterative algorithms are available to find the optimal stratum boundaries. The design can feature a user defined certainty stratum where all the units are sampled. Take-all and take-none strata can be included in the stratified design as they might lead to smaller sample sizes. The sample size calculations are based on the anticipated moments of the survey variable Y, given the stratification variable X. The package handles conditional distributions of Y given X that are either a heteroscedastic linear model, or a log-linear model. Stratum specific non-response can be accounted for in the design construction and in the sample size calculations.

Replication variance estimation under two-phase sampling

Jae Kwang Kim and Cindy Long Yu

Abstract

In two-phase sampling for stratification, the second-phase sample is selected by a stratified sample based on the information observed in the first-phase sample. We develop a replication-based bias adjusted variance estimator that extends the method of Kim, Navarro and Fuller (2006). The proposed method is also applicable when the first-phase sampling rate is not negligible and when second-phase sample selection is unequal probability Poisson sampling within each stratum. The proposed method can be extended to variance estimation for two-phase regression estimators. Results from a limited simulation study are presented.

Cost efficiency of repeated cluster surveys

Stanislav Kolenikov and Gustavo Angeles

Abstract

We analyze the statistical and economic efficiency of different designs of cluster surveys collected in two consecutive time periods, or waves. In an independent design, two cluster samples in two waves are taken independently from one another. In a cluster-panel design, the same clusters are used in both waves, but samples within clusters are taken independently in two time periods. In an observation-panel design, both clusters and observations are retained from one wave of data collection to another. By assuming a simple population structure, we derive design variances and costs of the surveys conducted according to these designs. We first consider a situation in which the interest lies in estimation of the change in the population mean between two time periods, and derive the optimal sample allocations for the three designs of interest. We then propose the utility maximization framework borrowed from microeconomics to illustrate a possible approach to the choice of the design that strives to optimize several variances simultaneously. Incorporating the contemporaneous means and their variances tends to shift the preferences from observation-panel towards simpler panel-cluster and independent designs if the panel mode of data collection is too expensive. We present numeric illustrations demonstrating how a survey designer may want to choose the efficient design given the population parameters and data collection cost.

On the efficiency of randomized probability proportional to size sampling

Paul Knottnerus

Abstract

This paper examines the efficiency of the Horvitz-Thompson estimator from a systematic probability proportional to size (PPS) sample drawn from a randomly ordered list. In particular, the efficiency is compared with that of an ordinary ratio estimator. The theoretical results are confirmed empirically with of a simulation study using Dutch data from the Producer Price Index.

Short Notes:

The use of estimating equations to perform a calibration on complex parameters

Éric Lesage

Abstract

In the calibration method proposed by Deville and Särndal (1992), the calibration equations take only exact estimates of auxiliary variable totals into account. This article examines other parameters besides totals for calibration. Parameters that are considered complex include the ratio, median or variance of auxiliary variables.