Survey Methodology
Archived Content
Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.
Release date: June 22, 2017
The journal Survey Methodology Volume 43, Number 1 (June 2017) contains the following 7 papers:
Waksberg Invited Paper Series
The promise and challenge of pushing respondents to the Web in mixed‑mode surveys
Web-push survey data collection that uses mail contact to request responses over the Internet, while withholding alternative answering modes until later in the implementation process, has developed rapidly over the past decade. This paper describes the reasons this innovative mixing of survey contact and response modes was needed, the primary ones being the declining effectiveness of voice telephone and slower than expected development of email/web only data collection methods. Historical and institutional barriers to mixing survey modes in this manner are also discussed. Essential research on the use of U.S. Postal address lists and the effects of aural and visual communication on survey measurement are then described followed by discussion of experimental efforts to create a viable web-push methodology as an alternative to voice telephone and mail response surveys. Multiple examples of current and anticipated web-push data collection uses are provided. This paper ends with a discussion of both the great promise and significant challenge presented by greater reliance on web-push survey methods.
Regular Papers
A layered perturbation method for the protection of tabular outputs
The protection of data confidentiality in tables of magnitude can become extremely difficult when working in a custom tabulation environment. A relatively simple solution consists of perturbing the underlying microdata beforehand, but the negative impact on the accuracy of aggregates can be too high. A perturbative method is proposed that aims to better balance the needs of data protection and data accuracy in such an environment. The method works by processing the data in each cell in layers, applying higher levels of perturbation for the largest values and little or no perturbation for the smallest ones. The method is primarily aimed at protecting personal data, which tend to be less skewed than business data.
State space time series modelling of the Dutch Labour Force Survey: Model selection and mean squared errors estimation
by Oksana Bollineni-Balabay, Jan van den Brakel and Franz Palm
Structural time series models are a powerful technique for variance reduction in the framework of small area estimation (SAE) based on repeatedly conducted surveys. Statistics Netherlands implemented a structural time series model to produce monthly figures about the labour force with the Dutch Labour Force Survey (DLFS). Such models, however, contain unknown hyperparameters that have to be estimated before the Kalman filter can be launched to estimate state variables of the model. This paper describes a simulation aimed at studying the properties of hyperparameter estimators in the model. Simulating distributions of the hyperparameter estimators under different model specifications complements standard model diagnostics for state space models. Uncertainty around the model hyperparameters is another major issue. To account for hyperparameter uncertainty in the mean squared errors (MSE) estimates of the DLFS, several estimation approaches known in the literature are considered in a simulation. Apart from the MSE bias comparison, this paper also provides insight into the variances and MSEs of the MSE estimators considered.
Bayesian predictive inference of a proportion under a two-fold small area model with heterogeneous correlations
by Danhyang Lee, Balgobin Nandram and Dalho Kim
We use a Bayesian method to infer about a finite population proportion when binary data are collected using a two-fold sample design from small areas. The two-fold sample design has a two-stage cluster sample design within each area. A former hierarchical Bayesian model assumes that for each area the first stage binary responses are independent Bernoulli distributions, and the probabilities have beta distributions which are parameterized by a mean and a correlation coefficient. The means vary with areas but the correlation is the same over areas. However, to gain some flexibility we have now extended this model to accommodate different correlations. The means and the correlations have independent beta distributions. We call the former model a homogeneous model and the new model a heterogeneous model. All hyperparameters have proper noninformative priors. An additional complexity is that some of the parameters are weakly identified making it difficult to use a standard Gibbs sampler for computation. So we have used unimodal constraints for the beta prior distributions and a blocked Gibbs sampler to perform the computation. We have compared the heterogeneous and homogeneous models using an illustrative example and simulation study. As expected, the two-fold model with heterogeneous correlations is preferred.
Sample allocation for efficient model-based small area estimation
by Mauno Keto and Erkki Pahkinen
We present research results on sample allocations for efficient model-based small area estimation in cases where the areas of interest coincide with the strata. Although model-assisted and model-based estimation methods are common in the production of small area statistics, utilization of the underlying model and estimation method are rarely included in the sample area allocation scheme. Therefore, we have developed a new model-based allocation named g1-allocation. For comparison, one recently developed model-assisted allocation is presented. These two allocations are based on an adjusted measure of homogeneity which is computed using an auxiliary variable and is an approximation of the intra-class correlation within areas. Five model-free area allocation solutions presented in the past are selected from the literature as reference allocations. Equal and proportional allocations need the number of areas and area-specific numbers of basic statistical units. The Neyman, Bankier and NLP (Non-Linear Programming) allocation need values for the study variable concerning area level parameters such as standard deviation, coefficient of variation or totals. In general, allocation methods can be classified according to the optimization criteria and use of auxiliary data. Statistical properties of the various methods are assessed through sample simulation experiments using real population register data. It can be concluded from simulation results that inclusion of the model and estimation method into the allocation method improves estimation results.
A mixed latent class Markov approach for estimating labour market mobility with multiple indicators and retrospective interrogation
by Francesca Bassi, Marcel Croon and Davide Vidotto
Measurement errors can induce bias in the estimation of transitions, leading to erroneous conclusions about labour market dynamics. Traditional literature on gross flows estimation is based on the assumption that measurement errors are uncorrelated over time. This assumption is not realistic in many contexts, because of survey design and data collection strategies. In this work, we use a model-based approach to correct observed gross flows from classification errors with latent class Markov models. We refer to data collected with the Italian Continuous Labour Force Survey, which is cross-sectional, quarterly, with a 2-2-2 rotating design. The questionnaire allows us to use multiple indicators of labour force conditions for each quarter: two collected in the first interview, and a third one collected one year later. Our approach provides a method to estimate labour market mobility, taking into account correlated errors and the rotating design of the survey. The best-fitting model is a mixed latent class Markov model with covariates affecting latent transitions and correlated errors among indicators; the mixture components are of mover-stayer type. The better fit of the mixture specification is due to more accurately estimated latent transitions.
Variance estimation in multi-phase calibration
by Noam Cohen, Dan Ben-Hur and Luisa Burck
The derivation of estimators in a multi-phase calibration process requires a sequential computation of estimators and calibrated weights of previous phases in order to obtain those of later ones. Already after two phases of calibration the estimators and their variances involve calibration factors from both phases and the formulae become cumbersome and uninformative. As a consequence the literature so far deals mainly with two phases while three phases or more are rarely being considered. The analysis in some cases is ad-hoc for a specific design and no comprehensive methodology for constructing calibrated estimators, and more challengingly, estimating their variances in three or more phases was formed. We provide a closed form formula for the variance of multi-phase calibrated estimators that holds for any number of phases. By specifying a new presentation of multi-phase calibrated weights it is possible to construct calibrated estimators that have the form of multi-variate regression estimators which enables a computation of a consistent estimator for their variance. This new variance estimator is not only general for any number of phases but also has some favorable characteristics. A comparison to other estimators in the special case of two-phase calibration and another independent study for three phases are presented.
- Date modified: