Survey Methodology
Release date: December 17, 2019
The journal Survey Methodology Volume 45, Number 3 (December 2019) contains eight papers and one short note.
Regular Papers
Estimation of level and change for unemployment using structural time series models
by Harm Jan Boonstra and Jan A. van den Brakel
Monthly estimates of provincial unemployment based on the Dutch Labour Force Survey (LFS) are obtained using time series models. The models account for rotation group bias and serial correlation due to the rotating panel design of the LFS. This paper compares two approaches of estimating structural time series models (STM). In the first approach STMs are expressed as state space models, fitted using a Kalman filter and smoother in a frequentist framework. As an alternative, these STMs are expressed as time series multilevel models in an hierarchical Bayesian framework, and estimated using a Gibbs sampler. Monthly unemployment estimates and standard errors based on these models are compared for the twelve provinces of the Netherlands. Pros and cons of the multilevel approach and state space approach are discussed.
Robust variance estimators for generalized regression estimators in cluster samples
by Timothy L. Kennel and Richard Valliant
Standard linearization estimators of the variance of the general regression estimator are often too small, leading to confidence intervals that do not cover at the desired rate. Hat matrix adjustments can be used in two-stage sampling that help remedy this problem. We present theory for several new variance estimators and compare them to standard estimators in a series of simulations. The proposed estimators correct negative biases and improve confidence interval coverage rates in a variety of situations that mirror ones that are met in practice.
A note on propensity score weighting method using paradata in survey sampling
by Seho Park, Jae Kwang Kim and Kimin Kim
Paradata is often collected during the survey process to monitor the quality of the survey response. One such paradata is a respondent behavior, which can be used to construct response models. The propensity score weight using the respondent behavior information can be applied to the final analysis to reduce the nonresponse bias. However, including the surrogate variable in the propensity score weighting does not always guarantee the efficiency gain. We show that the surrogate variable is useful only when it is correlated with the study variable. Results from a limited simulation study confirm the finding. A real data application using the Korean Workplace Panel Survey data is also presented.
Suggestion of confidence interval methods for the Cronbach alpha in application to complex survey data
by Jihnhee Yu, Ziqiang Chen, Kan Wang and Mine Tezal
We discuss a relevant inference for the alpha coefficient (Cronbach, 1951) - a popular ratio-type statistic for the covariances and variances in survey sampling including complex survey sampling with unequal selection probabilities. This study can help investigators who wish to evaluate various psychological or social instruments used in large surveys. For the survey data, we investigate workable confidence intervals by using two approaches: (1) the linearization method using the influence function and (2) the coverage-corrected bootstrap method. The linearization method provides adequate coverage rates with correlated ordinal values that many instruments consist of; however, this method may not be as good with some non-normal underlying distributions, e.g., a multi-lognormal distribution. We suggest that the coverage-corrected bootstrap method can be used as a complement to the linearization method, because the coverage-corrected bootstrap method is computer-intensive. Using the developed methods, we provide the confidence intervals for the alpha coefficient to assess various mental health instruments (Kessler 10, Kessler 6 and Sheehan Disability Scale) for different demographics using data from the National Comorbidity Survey Replication (NCS-R).
Cost optimal sampling for the integrated observation of different populations
by Piero Demetrio Falorsi, Paolo Righi and Pierre Lavallée
Social or economic studies often need to have a global view of society. For example, in agricultural studies, the characteristics of farms can be linked to the social activities of individuals. Hence, studies of a given phenomenon should be done by considering variables of interest referring to different target populations that are related to each other. In order to get an insight into an underlying phenomenon, the observations must be carried out in an integrated way, in which the units of a given population have to be observed jointly with related units of the other population. In the agricultural example, this means that a sample of rural households should be selected that have some relationship with the farm sample to be used for the study.
A grouping genetic algorithm for joint stratification and sample allocation designs
by Mervyn O’Luing, Steven Prestwich, and S. Armagan Tarim
Finding the optimal stratification and sample size in univariate and multivariate sample design is hard when the population frame is large. There are alternative ways of modelling and solving this problem, and one of the most natural uses genetic algorithms (GA) combined with the Bethel-Chromy evaluation algorithm. The GA iteratively searches for the minimum sample size necessary to meet precision constraints in partitionings of atomic strata created by the Cartesian product of auxiliary variables. We point out a drawback with classical GAs when applied to the grouping problem, and propose a new GA approach using “grouping” genetic operators instead of traditional operators. Experiments show a significant improvement in solution quality for similar computational effort.
“Optimal” calibration weights under unit nonresponse in survey sampling
by Per Gösta Andersson
High nonresponse is a very common problem in sample surveys today. In statistical terms we are worried about increased bias and variance of estimators for population quantities such as totals or means. Different methods have been suggested in order to compensate for this phenomenon. We can roughly divide them into imputation and calibration and it is the latter approach we will focus on here. A wide spectrum of possibilities is included in the class of calibration estimators. We explore linear calibration, where we suggest using a nonresponse version of the design-based optimal regression estimator. Comparisons are made between this estimator and a GREG type estimator. Distance measures play a very important part in the construction of calibration estimators. We show that an estimator of the average response propensity (probability) can be included in the “optimal” distance measure under nonresponse, which will help to reduce the bias of the resulting estimator. To illustrate empirically the theoretically derived results for the suggested estimators, a simulation study has been carried out. The population is called KYBOK and consists of clerical municipalities in Sweden, where the variables include financial as well as size measurements. The results are encouraging for the “optimal” estimator in combination with the estimated average response propensity, where the bias was reduced for most of the Poisson sampling cases in the study.
A method to correct for frame membership error in dual frame estimators
by Dong Lin, Zhaoce Liu and Lynne Stokes
Dual frame surveys are useful when no single frame with adequate coverage exists. However estimators from dual frame designs require knowledge of the frame memberships of each sampled unit. When this information is not available from the frame itself, it is often collected from the respondent. When respondents provide incorrect membership information, the resulting estimators of means or totals can be biased. A method for reducing this bias, using accurate membership information obtained about a subsample of respondents, is proposed. The properties of the new estimator are examined and compared to alternative estimators. The proposed estimator is applied to the data from the motivating example, which was a recreational angler survey, using an address frame and an incomplete fishing license frame.
Short note
On a new estimator for the variance of the ratio estimator with small sample corrections
by Paul Knottnerus and Sander Scholtus
The widely used formulas for the variance of the ratio estimator may lead to serious underestimates when the sample size is small; see Sukhatme (1954), Koop (1968), Rao (1969), and Cochran (1977, pages 163-164). In order to solve this classical problem, we propose in this paper new estimators for the variance and the mean square error of the ratio estimator that do not suffer from such a large negative bias. Similar estimation formulas can be derived for alternative ratio estimators as discussed in Tin (1965). We compare three mean square error estimators for the ratio estimator in a simulation study.
- Date modified: