Thursday, June 01, 2006
This blog wonders what is the potential of the ClustrMap system that is there in the left margin of the blog. A very useful application of the system is to survey how many people from how many countries have access to the internet and thus realize how many people are well-off enough to own computers that havw access to the internet.
A simple manner to conduct this survey is by subscribing a ClustrMap for a popular search engine, say google. Since there is a high probablity for an internet user to use the search engine to find information on the internet, a ClustrMap will thus be able to simply identify the number of people from each region by size of circle.
However, there are numerous limitations to this approach. The first and most important is the limitations of ClustrMap itself. The program may not be able keep track of the large amount of persons streaming into google, leading to a problem of breakdown of the system. Assuming this technical problem can be fixed, there are other intrinsic problems as well. First is there will be under-reporting in non-english speaking countries, especially China and Japan, who have native search engines and will not rely on google. In addition, it is not unforeseeable that severval different persons maybe sharing the same computer, leading to a under-reporting in the number of IPs, and thus number of users.
The most interesting question however, is what is the use of this survey. Will there be any use of such a survey? This perhaps can be seen as an alternative to government census where such census is not available or cannot be trusted. However, it is a very roundabout manner of getting information, considering that as of now the internet is not considered an essential good. Perhaps one day, the census will consist of some sort of electronic survey, such as the one discribed above, which is much cheaper than current surveys.
10:39 PM
