On a Modular Approach to the Design of Integrated Social Surveys

Abstract

This article considers a modular approach to the design of integrated social surveys. The approach consists of grouping variables into 'modules', each of which is then allocated to one or more 'instruments'. Each instrument is then administered to a random sample of population units, and each sample unit responds to all modules of the instrument. This approach offers a way of designing a system of integrated social surveys that balances the need to limit the cost and the need to obtain sufficient information. The allocation of the modules to instruments draws on the methodology of split questionnaire designs. The composition of the instruments, that is, how the modules are allocated to instruments, and the corresponding sample sizes are obtained as a solution to an optimisation problem. This optimisation involves minimisation of respondent burden and data collection cost, while respecting certain design constraints usually encountered in practice. These constraints may include, for example, the level of precision required and dependencies between the variables. We propose using a random search algorithm to find approximate optimal solutions to this problem. The algorithm is proved to fulfil conditions that ensure convergence to the global optimum and can also produce an efficient design for a split questionnaire.

Optimal Stratification and Allocation for the June Agricultural Survey

“…Application to more complex designs has not been considered in this research, but the model-assisted objective function could be extended to account for subsampling and other traits of complex designs. Application to more complicated designs requiring optimization of multiple samples may also be possible for paneled or split-questionnaire designs as in Ioannidis et al (2016).…”

Block Weighted Least Squares Estimation for Nonlinear Cost-based Split Questionnaire Design

“…Generally, these parameters can be determined according to practical considerations or operational data. Figure 1 demonstrates an example of a j with B 2 ¼ 0:1; B 1 ¼ 21; B 0 ¼ 10: Note that when B 2 ¼ 0; the cost function corresponds to the linear cost function (Ioannidis et al 2016).…”

“…With the increasing interest in SQD, the optimal design obtained by minimising a proper cost function for the survey, subject to specified reliability constraints, is an important aspect for more research. For example, Steel (2009, 2011) and Ioannidis et al (2016) introduced a linear cost function, which facilitates the search for the optimal design. However, the linear form of cost function may still be insufficient to accommodate a potentially complex relationship between the number of questions and the cost of the survey.…”

Modularization in an Era of Mobile Web

“…Raghunathan and Grizzle (1995) use a Bayesian approach for split questionnaire design, while Chipperfield and Steel (2009) focus on a design-based approach. Ioannidis et al (2016) propose an optimization involving minimization of respondent burden and data collection cost, while respecting certain design constrains usually encountered in practice such as the level of precision required and dependencies between variables. Difficulties in split questionnaire design arise in the timing and frequency of different surveys, or how to implement rotating panels overlapping in time.…”

“…Raghunathan and Grizzle (1995) use a Bayesian approach for split questionnaire design, while Chipperfield and Steel (2009) focus on a design-based approach. Ioannidis et al (2016) propose…”