Online Program

Embedded Experiment for Non-Response Follow-Up Methods of Electronic Questionnaire Collection

Keywords: Electronic Collection,Follow-Up,Experimental Design,Paradata

At Statistics Canada, many business surveys use mail questionnaires for initial data gathering and telephone follow-up to collect data from respondents who have not returned the questionnaires after a pre-specified period. Currently, Statistics Canada is undertaking a general restructuring of its business statistics programs. One of its goals is to make electronic data collection the principal mode of collection for business surveys. Until now, follow-up methods used for electronic questionnaire were based on paper questionnaire collection methods. International and Statistics Canada experiences have shown that electronic respondents have different follow-up patterns than paper respondents. In order to establish a standard collection follow-up strategy for electronic questionnaire, an experiment embedded into the survey was conducted to compare different non-response follow-up methods (combining phone, and e-mail reminder). Analysis of paradata outputs such as response rates, time spent on follow-up and time of response would permit us to compare the different methods. Those paradata are generated by the Blaise system that records the history of all calls made to enterprises throughout the collection cycle. This document presents the results of this experiment and other ongoing investigations on electronic data collection. Through these results, we aims to identify the impact of using only e-mail reminders for follow-up. Furthermore, we measure the importance of having the first contact via telephone versus e-mail reminder. We also identify and discuss the most efficient follow-up strategy that yields the best response rates. Finally, we establish standard collection follow-up strategies for multi-mode collection involving electronic questionnaires.