[01] Ang Khay Wee, Department of Statistics and Applied Probability, National University of Singapore, Singapore, Singapore. [02] Carol Anne Hargreaves, Department of Statistics and Applied Probability, National University of Singapore, Singapore, Singapore.

This survey design is motivated by the need to address the consequences such as non-response, high respondent burden and poor data quality, that lengthy surveys are often associated with. Responsive split design is a survey design technique that directly addresses these consequences. The objective of this study was to reduce the respondent burden and to increase the response rate by using the Split Questionnaire Design (SQD). We leveraged on the idea of issuing only relevant surveys to each respondent as this directly reduces the connotation of being probed with irrelevant questions which is regarded as a primary reason for non-response. To achieve this, we developed a responsive design by utilising the prior information that we collected as part of the questionnaire, and then created decision rules for the administration of the relevant micro-surveys. The results that we have attained are promising in terms of the trade-off between precision and responsiveness. The responsive design offered the advantage of being more responsive to the tourist and was able to detect rare expenditures. This suggests that the responsive design provides improvements in issues regarding missing and rare events. Besides meeting the objectives, we also demonstrated that a secondary advantage of responsive split questionnaires also brings about significant cost savings to the survey organisation.

Responsive Split Questionnaire Design, Split Questionnaire Design, Machine Learning, Tourism

[01] Adiguzel F. and Wedel M. (2008). Split Questionnaire Design for Massive Surveys. Journal of Marketing Research, 45 (5), 608-612. [02] Ali, M., Kauermann, G. (2021). A split questionnaire survey design in the context of statistical matching. Stat Methods and Applications. https://doi.org/10.1007/s10260-020-00554-2 [03] Breidt J. and Opsomer J. (2017). Model-Assisted Survey Estimation with Modern Prediction Techniques. Statistical Science, 190-205. [04] Chun A. Y., Heeringa G., Schouten B. (2018). Responsive and Adaptive Design for Survey Optimization. Journal of Official Statistics, pp 581-597. doi: 10.2478/jos-2018-0028. [05] Chipperfield, J. O., Barr, M. L. and Steel, D. G., (2018). Split Questionnaire Designs: collecting only the data that you need through MCAR and MAR designs. Journal of Applied Statistics, 1465-1475. doi: 10.1080/02664763.2017.1375085. [06] Christian P. Robert, G. C. (2004). Monte Carlo Statistical Methods. Springer. doi: 10.1007/978-1-4757-4145-2. [07] Cochran, W. G. (1977). Sampling techniques 3rd Edition. New York: Wiley. [08] Early K, Mankoff J, Fienberg S. (2017). Dynamic Question Ordering in Online Surveys. Journal of Official Statistics, 625-657. doi: 10.1515/jos-2017-0030. [09] Gonzalez, M. (2012). The use of responsive split questionnaires in a panel survey. [10] Groves M. and Heeringa G. (2006). Responsive design for household surveys: tools for actively controlling survey errors and costs. Journal of the Royal Statistical Society Series A: Stiatistics in Society, 169 (3), pp. 439-457. doi: 10.1111/j.1467-985X.2006.00423.x. [11] Hundepool, A. (2012). Statistical Disclosure Control - Wiley Series in Survey Methodology. John Wiley & Sons Inc. [12] Murphy J., Biemer P. and Berry C. (2018). Transitioning a Survey to Self-Administration using Adaptive, Responsive, and Tailored (ART) Design Principles and Data Visualization. Journal of Official Statistics, Vol. 34, 625-648. doi: 10.2478/jos-2018-0030. [13] Rolstad, S. (Dec, 2011). Response Burden and Questionnaire Length: Is Shorter Better? A Review and Meta-analysis. Value in Health, 1101-1108. doi: https://doi.org/10.1016/j.jval.2011.06.003 [14] Thompson, D. G. (1952). A Generalization of Sampling Without Replacement From a Finite Universe. American Statistical Association, 663-685. [15] Wagner, J. R. (2008). Adaptive survey design to reduce nonresponse bias. PhD Thesis University of Michigan. [16] Wallis, William H. Kruskal and W. Allen. (1952). Use of Ranks in One-Criterion Variance Analysis. Journal of the American Statistical Association, 583-621. doi: 10.2307/2280779.