It’s no secret that, since Xpiry Day, some businesses have been buying hardware that officially ships with Vista, but exploiting a loophole to legally ‘downgrade’ to XP. Now we have the first specific numerical claim, with a report that 35% of firms are doing this.

The data comes from the InfoWorld site’s Windows Sentinel project. This is a scheme were participating users install a piece of tracking software on their machines which monitors performance and alerts users to possible problems. Of course, the real point of the project is that it provides aggregated details on how well various versions of Windows are performing.

You may remember this project from a recent promotion in which people with the software installed could get a prediction of whether or not their machines would be able to cope with Windows 7.

The people behind the project have run some analysis taking advantage of the fact that Sentinel gathers data about a computer’s original manufacturer as well as the edition of Windows that is running. Based on these details, they believe 35% of computers they’ve scanned which should be running Vista are running another system.

The problem with this headline statistic is that there are plenty of caveats. For one thing, it’s not entirely clear whether they’ve been able to distinguish between business and home users in these statistics. They’ve also either don’t know or won’t say what proportion are actually downgrading to XP: the 35% figure also includes people who’ve replaced Vista with a Microsoft Server product or even Linux.

The other potential limitation is the sample size itself. There are only 3,000 participants in the entire project, so the number who ‘should’ be running Vista is somewhat lower. That’s not necessarily a reason to discount the results as mathematically we can still be confident the ‘true’ figure is somewhere in the 30-40% range based on this sample size.

The problem is that this mathematical reliability assumes you are working from a representative sample. By only taking results from participants in the project, there’s a clear bias towards people with a particular interest in the performance of their PCs; at the very least, you’ve completely ignored the “plug it in and hope for the best” sector of the market.

These are still the first publicly available XP downgrade figures based on quantifiable facts rather than rumour. But they don’t really tell us much other than some, but not most, Vista-machine buyers are downgrading to XP.

Related:

Entry was posted on Monday, August 18th, 2008 at 10:47 am under Microsoft, Upgrading, Vista, xp.  Print This Post

Read more entries by John Lister.
Stumble It!