Web and mixed-mode data collection in National Statistics 1
|Convenor||Mrs Karen Blanke (FSO Germany )|
|Coordinator 1||Mrs Annemieke Luiten (Statistics Netherlands)|
While the ESSnet project “Data Collection in Social Surveys Using multiple Modes” (DCSS) project is coming to an end, methodological discussion is ongoing. Many countries in the ESS plan to introduce web as part of their mixed-mode designs for social surveys . Based on the findings of the project there seems to be no fundamental reason to advise against this strategy. The presentation discusses several advantages and disadvantages of introducing web data collection. Based on lessons learnt from conducting the project (meeting reality), issues such as unimode vs specified approaches for questionnaire design, standardization and longitudinal perspectives will be presented.
The Office for National Statistics is investing in the introduction of an online mode to its existing Social Surveys and an online Census in 2021. This is an important ongoing area of work; previous experience has highlighted the need for a radical research programme to support successful introduction of online data collection. This presentation sets out the plans for ongoing research and the context within which this work is being taken forward. It shares the research findings and challenges faced to date and those on the horizon. We invite discussion and encourage others to share their experiences and recommendations.
The American Community Survey (ACS) successfully employs a sequential multi-mode collection strategy that includes web, mailout/mailback, telephone, and face-to-face interviewing. This presentation summarizes results from several analyses of survey data as well as recent field experiments to refine our methods and answer the following questions: Can reminder email messages effectively encourage respondents to finish partially completed survey responses? What modifications to the mail materials encourage response by web or mail? What are the characteristics of those who complete the survey using mobile devices? How does the use of a web response option impact response groups with
The Statistical Research Institute has studied the design of the web questionnaire of the 2015 Census using an eye-tracker. If a non response or error was detected in web survey, respondents received the “edit” message with a pop up window, and the respondents were required to fill in the missing information or correct information. However, they were easily fatigued and tended to answer less accurately when the respondents received too many messages. For this sort of “response fatigue” led to satisficing errors. Thus, we should consider what the appropriate number of edit messages is to be in web survey