Seemed like a slightly odd focus for study for me, given that advertisers do what they do and website visitors click or don't click. Clickthrough traffic is monitored constantly and at the end of the day, if no one clicked, then advertisers wouldn't do it. But then, I am oversimplifying things here, and this was an academic study...
There were some interesting outcomes - while not everyone looked at the banner ads, in tests where ads were present on some pages and not on others, there was an increase in attention compared to tests where ads were consistently placed. Also ad recognition post-test was higher among those who'd been exposed during the test.
The conclusion from the blog post which summarised this study:
The lessons for web advertisers are clear: don't advertise on every page, vary ad location, and make sure the ad topic is congruent with the web-site content.As I work in the public sector my interest in this study wasn't so much about the impact on advertising placement. There are lessons to be learned though about placing important content in locations on the page that are commonly associated with advertising banners.
Are we really blind to Internet banners? - blog post by Christian Jarrett on BPS Research Digest