ClearType improves our accuracy at recognizing words

Most people agree that ClearType makes text look nicer, especially when tuned. But does it help us read more efficiently? This is a difficult question because adult readers can read at a very fast rate – good readers can read 250-300 words per minute, which is a word every 200-240 milliseconds on average. Measuring something that takes less than a quarter of a second is a challenge.

Tachistoscopic Lexical Decision is a technique that has been used extensively to measure word recognition. Tachistoscopes were a mechanical device for presenting materials briefly, but have been replaced by computers. Readers are briefly shown either a regular word in their native language (e.g. distant) or a pseudoword (e.g. tadints), which is not a word in their native language. The reader’s task is to say if they saw a word or pseudoword. The advantage to this methodology it only requires the reader to complete a very simple task of pressing one button for words or another for pseudowords. This allows more accurate timing for word recognition.

The below animated image demonstrates what the reader sees during this task. The word and pseudoword here are appearing on screen for a longer period than in a typical study, but special software is needed to create shorter presentations.

A team led by Professors Lee Gugerty and Rick Tyrrell at Clemson University studied the effects of ClearType on the tachistoscopic lexical decision task. They asked 24 readers to respond to 280 words and 280 pseudowords. Half of the words were presented with ClearType rendering and half with the regular black & white rendering. They found that readers were statistically reliably more accurate at recognizing words when words were rendered with ClearType.

After correcting for guessing, this translates to a 17% accuracy improvement at recognizing words with ClearType than with black & white. This is a large difference in word recognition performance with a task that simulates the real world situation of looking at words for very brief periods of time.

Kevin Larson

Gugerty, L., Tyrrell, R. A., Aten, T. R. & Edmonds, K. A. (2004). The effects of subpixel addressing on users’ performance and preferences during reading-related tasks. ACM Transactions on Applied Perception, 1(2), 81-101.

Edit: Update Image Reference

Comments (14)

  1. Anonymous says:

    That bar graph is kind of misleading, don’t you think? Looks like a >50% improvement if you just look at the bars..

  2. Kevin Larson says:

    It wasn’t my intention to misrepresent the magnitude of effect, but rather was trying to zoom in on the two bars to make it clear that the error bars for ClearType and black & white do not overlap. A full scale would show lexical decision accuracy ranging from a low of 50%, or chance response, to a maximum of 100% accurate.

  3. IEBlog says:

    Hi. My name is Bill Hill, and I’m one of the original inventors of ClearType. My job at Microsoft is…

  4. I have posted before on this issue, several times, in fact. And so have many others.

    But now there…

  5. CodeObsessed says:

    In particular, I don’t. Lots of times other developers come over to me and look confused as they see

  6. fontblog says:

    In two earlier posts I talked about studies that showed that word recognition is 17% more accurate and

  7. IEBlog says:

    For Beta 1, we discussed some of the technical improvements (like domain highlighting, multi-line paste,

  8. fontblog : ClearType improves our accuracy at recognizing words

  9.   Internet Explorer 8 Beta 1에서 주소 표시줄에 대해서 실시한 몇가지  기술적인 안내 (도메인 하이라이트 기능, 여러 행 붙이기 기능, 클릭

  10. John Doe says:

    Again, it goes without saying that this is relative to black & white (e.g. no smooth edges in Windows), not gray scale anti-aliasing (e.g. "Standard" smooth edges in Windows).