Xyologic, a company specializing in app discovery and search, has released the results of a study into app preferences among iPhone and Android users. The study compared the top 150 downloads from the Android App Market in November with those from the iPhone App Store. They concluded that “iPhone is for games, Android is for apps.”

Xyologic’s study found that 85 of the top 150 downloads in the Android App Market were apps, and the remaining 65 were games. Conversely, only 50 of the top 150 downloads in the iPhone store were apps, while 100 were games. Both stores showed a significant increase in game downloads over last year, when 79 of the top 150 in the iPhone store and only 34 in the Android App Market were games.

At first glance, this data seems to paint an interesting picture of phone usage between iOS and Android users. There is, however, a fatal flaw in Xyologic’s methodology: they compared the top 150 Android downloads overall to the top 150 free iPhone apps. In other words, the iOS data is skewed. The data excludes two immensely important categories of apps: iPad apps, and paid apps. Xyologic compared the top 150 Android downloads overall to only one segment of App Store downloads.

In response to a flurry of comments pointing this flaw out, Xyologic added an update to the blog post where they originally published the results:

Please note that all the app [sic] in the Android Top 150 are free apps. Hence this is a comparison of free US Android Market apps with free US Apple App Store iPhone apps.

This alleviates the problem somewhat, but not entirely. It still isn’t a fair comparison. They’ve still compared the top overall Android downloads to a narrow subset of iOS downloads. The fact that all the Android apps in question happen to be free is effectively a coincidence: it may just mean that Android users are less willing to pay for apps.

While a proper comparison of the top 150 downloads in the App Market versus the App Store would certainly be interesting – and may even come to a similar conclusion – that is not what Xyologic has done. By narrowing the App Store results but not the Android results, they’ve skewed the data and tainted the results.