The Technology of Text – IEEE Spectrum

Table of Contents

Having gone to such lengths to make computer text reader-friendly, how do we know it’s working? When software developers try to gauge the effectiveness of a ­programming upgrade, they typically look at whether or not users can accomplish tasks they couldn’t do previously. Because users are already comfortable with the basic task of reading, we instead try to measure things like reading speed and comprehension.

Peer-reviewed studies have consistently found that using ClearType boosts reading performance compared with other text-rendering systems. In a 2004 study, for instance, Lee Gugerty, a psychology professor at Clemson University, in South Carolina, measured a 17 percent improvement in word recognition accuracy with ClearType. Gugerty’s group also showed, in a sentence comprehension study, that ClearType boosted reading speed by 5 ­percent and comprehension by 2 ­percent. Those results were unusual because, typically, any gain in reading speed decreases comprehension.

Similarly, in a study published last year, psychologist Andrew Dillon at the University of Texas at Austin found that when subjects were asked to scan a spreadsheet and pick out certain information, they did those tasks 7 percent faster with ClearType.

A good way to measure a font’s readability is to conduct a visual acuity exam, which is similar to the eye test you have to pass to get a driver’s license. In one such study, James Sheedy, while he was an optometry professor at Ohio State University, in Columbus, compared various computer fonts to see which was the most readable. He concluded that Verdana is more legible than two other popular fonts, Georgia and Arial, and in turn Georgia and Arial are more legible than Times New Roman, which had been designed for The Times of London back in 1931.

Sheedy’s results confirmed what we already believed to be true about onscreen legibility: among these four typefaces, Verdana, which was specifically developed for computers, has the largest lowercase letter heights and the most generous spacing inside and between letters, while Times New Roman has the smallest lowercases and the least generous spacing.

But we’ve also found that you can’t always judge how successful a reading technology is just by looking at reading speed and comprehension. In one study, we had subjects peruse documents with good page layouts and bad layouts. In the bad layout, ­readers had to jump over an image to keep reading across the line, while in the good layout, the image didn’t interfere with the text. While readers said they preferred the nicer layout, they exhibited no differences in reading speed or comprehension.

So we’ve started developing new methods for measuring onscreen reading. One promising area involves looking at how emotions affect behavior. For instance, when people receive a small gift, their mood improves, and they perform better on cognitive tasks. We wondered if the “pleasure” of reading a well-­rendered screen of text would also enhance performance. Amazingly, it does. [See sidebar, “Mastering the Candle Task.”]

More recently, we’ve found physical evidence of these emotional effects by measuring facial muscles associated with smiling and frowning. We reran the page layout test in which we’d previously failed to detect any differences in reading speed or comprehension; this time, though, we found that people frowned more when reading the poor page layout. If bad layouts are affecting their moods, even subtly, that could in turn decrease their overall productivity.

In ongoing research sponsored by Microsoft, Keith Rayner at the University of Massachusetts, Amherst, is studying the impact of ClearType and other technologies on eye movements. When we read, we perceive that our eyes move smoothly across the line of text, but they actually are making jumps from word to word—fixating on a word for 250 to 300 ms, then making a quick movement to the next word. [For more on how we read, see sidebar, “Mystery of the Scrambled Words.”]

As type designers and engineers continue to develop new and better technologies to render onscreen text, and psychologists continue to study their effectiveness, we eventually hope to make it as easy and comfortable to read from the computer screen as it is from the page. Reaching such a goal will not only make us more productive but also help preserve literacy and spread the world’s knowledge. It might even save a few trees, too.

Solution to the Candle Task

Empty the tacks from their box, and then tack the empty box to the corkboard. Place the candle inside the box and light it; the wax will stay in the box.

About the Author

KEVIN LARSON holds a Ph.D. in cognitive psychology and is a member of the Microsoft Advanced Reading Technologies group, based in Redmond, Wash.

To Probe Further

More about ClearType and other typography-related endeavors at Microsoft is available at http://www.microsoft.com/typography.

A December 2005 interview with Microsoft type gurus Bill Hill and Geraldine Wade (and a cameo appearance by the author) is at http://channel9.msdn.com/Showpost.aspx?postid=146749.

For more in-depth discussions, see the following:

Kevin Larson, “The Science of Word Recognition,” July 2004, http://www.microsoft.com/typography/ctfonts/WordRecognition.aspx.

Keith Rayner et al., “Raeding Wrods With Jubmled Lettres; There Is a Cost,” Psychological Science, Vol. 17, no. 3, March 2006, pp. 192–93.

Karen Cheng, Designing Type (Yale University Press, 2006).

Andrew Dillon et al., “Visual Search and Reading Tasks Using ClearType and Regular Displays,” in SIGCHI Conference on Human Factors in Computing Systems, ACM Press, 2006, pp. 503–11.

Lee Gugerty et al., “Sub-pixel Addressing Improves Performance,” ACM Transactions on Applied Perception, Vol. 1, no. 2, 2004, pp. 81–101.

Kevin Larson et al., “Measuring the Aesthetics of Reading,” in People and Computers XX: Proceedings of HCI 2006, Vol. 1, British Computer Society.

James E. Sheedy et al., “Text Legibility and the Letter Superiority Effect,” Human Factors, Vol. 47, no. 4, Winter 2005, pp. 797–815.