Recently a statistic class at Iowa State University did a project where they analyzed the AdSense data from AdMoolah. I thought this was a great idea, because it was the type of analysis I had in mind when I first started the site. They did things like analyze how factors such as categories, language and PageRank correlate to earnings. Check out the blog post.
ive always known myself to be very analytical. I love analyzing statistics thats why i love adsense and google analytics and making money!~
I think this was a bit different. We all have to "analyze" statistics but they went at it from a very "unemotional", disconnected viewpoint, and applied equations to the data rather than just eyeballing it, which I suspect most of us do. What I find interesting is that they don't have much in-depth knowledge. They concluded that seasons and PR have the greatest effect on earnings. I would never think of that (PR; seasons are obvious) because PR doesn't always translate to more visitors. Maybe something else is at play - higher PR = higher eCPM, higher PR = higher CPC...
Statistic is all about the average site. So, on average, a higher PR means more traffic. And on average, more traffic means more clicks. And on average, more clicks means more earnings. So that's why there is a correlation between PR and earnings.
That was my initial thought too, but they have access (I would hope) to traffic stats and would take that into account. And, if that was the case, I would think they could distinguish between PR/traffic or at least report that both would lead to an increase in revenue. The fact that nothing is said of traffic makes me wonder.