Friday, February 10, 2006
This study does absolutely nothing to help clear things up. Testing an unpatched version of IE versus any version of Firefox, which is almost by definition patched, is completely useless. Anyone who is using such a 'virgin' IE release probably isn't net-savvy enough to download Firefox, or even type in the correct URL. So the comparison is utterly invalid. This study would not have been more difficult to perform in a meaningful fashion: not to have done so reveals either extreme bias without the cleverness to even vaguely mask it, a failure to grasp experimental design principles in the first place.