Paper authored by Johan Erikson, Gustav Haraldsen and Ger Snijkers.
"The data collection process of statistical surveys is facing many challenges in the 21st century. In the information society, surveys and information are everywhere. The demand for timely information is increasing. On the other hand, the relationship between a survey organisation and its respondents has changed, due to e.g. new technologies and increasing focus on response burden. Motivating motivate people and businesses to respond to surveys needs new approaches. Response rates are declining for both telephone and self-administered surveys. Coverage is becoming a larger problem in telephone surveys as fewer people have listed telephone numbers. Enterprises have reduced the size of their organisations and are more and more reluctant to spend time on surveys, especially in times of economic crisis. This calls for new data collection strategies.
In response to these challenges statistical offices have implemented data collection strategies which are based on the use of register data and the implementation of mixed-mode surveys (often with web surveys as a primary mode). Also, innovation programs have been set up to investigate new opportunities brought about by new technologies. Both from a methodological and a management point of view this brings about new challenges to data collection. Methodological challenges deal with designs aimed at getting and combining quality data, i.e. accurate, complete and timely data from various sources. Management challenges deal with the planning of resources and money, as well as the controlling and monitoring of data collection processes using process and quality indicators aimed at increasing the likelihood of meeting the specified results and reducing risks. This also involves the availability and accessibility of existing large data sets. In addition, surveys no longer can be organised as stove pipes, but need to be coordinated with regard to sampling, questionnaire and communication strategy design."
This was the final paper of the seminar, and provoked much discussion. Here we encourage members of the data collection community to debate the issues raised.......
In their recent book, Big Data, Viktor Mayer-Schönberger and Kenneth Cukier suggest that the science of statistics as we know it is on the cusp of change: our reliance on sampling, and therefore the management of probabilities and confidence that are so c
The paper notes that new and large data sources, especially digital data extracted from the Internet, are cheap, fast, often internationally-focused and can be specifically directed towards the needs of clients. Users can therefore be expected to begin a