In 1997, John Morkes' and Jakob Nielsen's studied how people read on the web and concluded that writing for the web should be scannable, concise, and objective. This is the most often quoted study on the topic. For example, Allyson Troffer (2000) repeats the conclusions and calls for writing that is concise and scannable by using bulleted lists, chunked text, and short topics.
My own paper, "Reading on the Web" (2001) relied heavily on the Morkes & Nielsen study. Using their study, I wrote about applying the conclusions to teaching children how to read online.
Since then, though, I've been studying how to write hypertext and Nielsen's comment is troublesome: "People rarely read Web pages word by word; instead, they scan the page, picking out individual words and sentences."
There is, at least, anecdotal evidence that people read entire sections of text. For example, college students, who post and read daily blogs (web logs), report that they read every word of the blogs. Derek M Powazek says, "If you don't believe me, believe this: The number one use of the internet is email. This is proved in study after study. The Pew Internet Project has found that, of the 111 million Americans that have gone online, 95 percent said that they've sent email. 54 percent said they do it daily. Now, if it's so hard to read on a computer, why is reading and writing text the number one thing that happens on line?"
Or, Powazek says it's ironic to attend a computer conference and see someone "standing in front of a room of people who read and write all day, every day, on computer" and say that most people don't read online.
If you accept that most people only scan online documents, then what are the implications for writing? Why would I write any more than I needed to? Bulleted lists of simple phrases would be the logical result. This dichotomy of writing/reading piqued my interest in the topic of reading on the web. How do you write, if the reader only scans?
Morkes & Nielsen's study, which privileges form over content, seemed inadequate. I decided to review Morkes' and Nielsen's study to see what could be learned from a critical look at this seminal research.
I looked at texts used, the questions asked of readers and the conclusions drawn. Issues which arose led me to question the methodology. This paper critiques the study in four ways. First, I question the quality of the writing in the test-texts because they don't follow accepted guidelines for good writing. Second, the task asked of readers are limited mostly to locating information, and gives little regard to how the information is used by the reader. Third, the conclusion drawn was that when writing was scannable, concise and objective that usability would increase. Morkes & Nielsen tested this hypothesis by rewriting a text about the Java language. While usability was increased dramatically, the rewritten text still has major flaws, indicating that when the guidelines (scannable, concise, objective) are followed clear writing may or may not occur.
Finally, the methodology of testing sets of text, in order to find out how people read on the web, seems questionable. I'll conclude with suggestions on the parameters that should be considered in future studies of how people read on the web. Implications for writing for the web are also discussed.
