Main

 
Study Overview: Critique of "How Users Read on the Web"

Study Overview

Morkes & Nielsen's study was conducted in three stages, with the first two stages exploratory and qualitative, and the third study quantitative.

In the first study, both technical and non-technical readers were taken to the homepage of a website and asked to find specific information on that site. For example, at the Restaurants & Institutions magazine website, readers were asked to find information about a Las Vegas restaurant run by chef Charlie Trotter. The study concludes that while searching for information, readers wanted to use a search box, scanning was the norm, and the organization of the website was crucial to how easily they found the information.

In the second study, mostly non-technical readers visited three preselected websites and performed tasks that required reading and answering questions about the website. While performing the tasks, readers made a running commentary on the experience. Then the reader answered questions about the site.

"Standard questions for each site included these:

  • "What would you say is the primary purpose of the site?"

  • "How would you describe the site's style of writing?"

  • "How do you like the way it is written?"

  • "How could the writing in this website be improved?"

  • "How much do you like this site? Why?

  • "Do you have any advice for the writer or designer of this website?"

  • "Think back to the site you saw just before this one. Of the two sites, which did you like better? Why?

From the results, Morkes/Nielsen conclude that simple, informal writing is preferred, credibility is an important concern, outbound links can increase credibility, humor should be used with caution, readers want the information quickly, text should be scannable and concise, summaries are helpful, inverted pyramid style helps focus a text, hypertext is popular, and graphics and text should complement each other.

In the third test, Morkes & Nielsen used an isolated website about traveling in Nebraska, written in a promotional style, and rewrote the site in four test styles concise, scannable, objective, and combined (concise, objective, scannable). The original website and test sites are available online. The appendix section here reproduces the "Attractions" page from each set of webpages.

  1. Promotional writing "contained exaggeration, subjective claims, and boasting."

  2. Concise writing used a promotional style, but reduced word count and sometimes rewrote in an inverted pyramid style.

  3. Scannable writing was also in a promotional style, but "was written to encourage scanning, or skimming, of the text for information of interest. This version used bulleted lists, boldface text to highlight keywords, photo captions, shorter sections of text, and more headings." Objective writing changed the style to informational or factual: "It presented information without exaggeration, subjective claims, or boasting.

  4. Combined writing "had shorter word count, was marked up for scannability, and was stripped of marketese."

The study measured time spent on a task (two search tasks, one judgment task), task errors, memory (recognition and recall), time to recall of website's structure (draw a sitemap), and subjective satisfaction (questionnaire asking about quality, ease of use, likeability, affect on reader).

The major finding was that scannable, concise, and objective writing results in better scores on every measure.

Morkes & Nielsen then used these guidelines to rewrite a section from Sun's website about Java language. They conclude that scannable, concise, objective writing increased usability by 154%.