In my posting five days ago about the EyeTrack07 survey by the Poynter Institute, News readers get more from web than print, I took its findings at face value. But two commentators were more sceptical and have raised some questions about methodology that casts doubts on the results.
Like me Andrew Grant-Adamson was taken with the claim that a great majority of online readers were more likely than newsprint readers to read the text of a story. But this didn't ring true to him and while he was thinking about it he came across a comment by Neil Sanderson who also found the analysis unconvincing, as he pointed out on the Poynter site:
"I am wondering how the online stories compared in length to those in broadsheet and tabloid papers. Perhaps they were shorter, due to the online practice of "chunking" - breaking information into more digestible pieces. On the other hand, I've seen plenty of online stories that were longer than their print counterparts. This can happen when a story originating in a newspaper is enhanced and updated for the online edition where space is virtually unlimited. Could story length have affected the amount of each story that was read?"
This elicited a reply from Poynter's Rick Edmonds which said: "Good question. During some editing/deeper digging this week, we have found that short stories were more frequent in online and that likely had some influence on the overall result. However if you back short stories out and analyse results only for medium and long stories, those are read, once selected, as thoroughly online as in print." Sanderson - who also took the trouble to post to my site - commented: "So, rather than saying that online readers read more, Poynter is now saying they read as thoroughly as in print. That's a big difference."
He added: "And describing stories as 'medium' or 'long' really doesn't solve the problem. The percentage of each story read ought to be judged against the length (word count) of each story. The story-length issue isn't just important for comparing online versus print reading patterns, but also for comparing broadsheet versus tabloid."
So, says Grant-Adamson, "it looks as if the Poynter people themselves fell into trap of going for the quick headline and having to pull back later. There is, I think, a lot of importance and value in this new study but it is probably better to wait for the more considered analysis later in the year."
