Читать книгу Handbook of Web Surveys - Jelke Bethlehem - Страница 18

In October 2004, Statistics Netherlands started a pilot to find out whether a CASI approach could be used to collect data for yearly production statistics.

One of the approaches tested is denoted by Electronic Data Reporting (EDR). It was a system for responding companies to manage interviewing programs (generated by the Blaise system) on their own computers. The EDR software was sent to respondents on CD‐ROM, or respondents could download the software from the Internet.

After the installation of the software, new survey interviews could be sent to respondents by e‐mail. These electronic questionnaires were automatically imported in the EDR environment. A simple click would start the interview. After offline completion of the interview, the entered data were automatically encrypted and sent to Statistics Netherlands.

The pilot made clear that downloading the software was feasible. It should be preferred over sending a CD‐ROM because it was simpler to manage and less expensive, too. Some companies experienced problems with downloading and installing the software, because security settings of their computer systems and networks prevented them of doing so. User‐friendliness and ease of navigation turned out to be important issues for respondents.

For more information about this pilot, see Snijkers, Tonglet, and Onat (2004, 2005).

This form of CASI also has disadvantages. It requires respondents to have computer skills. They should be able to download and run the interviewing program. Couper, Blair, and Triplett (1999) also note that problems may be caused by that fact that different users may have different operating systems on their computers or different versions of the same operating system. This may require different versions of the interviewing program, and it must be known in advance which operating system a respondent has. Moreover, the size of an executable file may be substantial, which may complicate sending it by e‐mail.

E‐mail surveys had the advantages of speed and low costs. Compared with CAI they had the disadvantages of a poor user interface and lack of adequate editing and navigation facilities. An e‐mail questionnaire was just a paper questionnaire in an e‐mail. The Internet became more interesting for survey data collection after HTML 2.0 was introduced in 1995. HTML stands for Hypertext Markup Language. It is the markup language for web pages. The first version of HTML was developed by Tim Berners‐Lee in 1991. Version 2 of HTML included support for forms. This made it possible to transfer data from a user to the web server. Web pages could contain questions, and the answers could be collected by the server. Example 1.5 shows some applicative aspects of the HTML questions.