next up previous
Next: Cognitive delay Up: User Delay Models from Previous: User Delay Models from

Perceptual delay

Psychological studies [4] have shown that humans read through discrete eye movements, which consist of fixations and saccades. Fixations are the state in which the eyes are focused on a object statically. Saccades represent rapid eye movements from one fixation location to another. Information is absorbed only during fixation, which lasts from $60$ to $700ms$. A saccade takes about $30ms$.

We categorize modern computer display information into two categories: graphical and textual. Graphical information is presented through graphical objects which consist of 1) window widgets, such as buttons, radiobuttons, menus, $etc.$, and 2) pictures or figures. To simplify delay estimation, we assume each graphical object requires one separate fixation and each fixation lasts for at least $60ms$. Therefore, for each graphical object, a delay of $90ms$ (one minimal fixation plus one saccade) is added.

Text can be regarded as a special kind of graphical object. Psychological studies [4] have shown that college students can raud (read with comprehension) at a typical rate of 300 standard-length words per minute or five per second. The standard-length word was assumed to have six characters. For a text with $n$ words, the typical time for rauding is

\begin{displaymath}
T= \frac{n{\times}cpw}{Wpm\times6}=\frac{n{\times}cpw}{30}
(s)
\end{displaymath}

where $cpw$ is the average number of characters per word and $Wpm$ is the rauding rate in terms of the number of standard-length words per minute. If a text is associated with a graphical object, the reading time is used as the fixation length for the graphical object and there is no separate delay for text reading.
next up previous
Next: Cognitive delay Up: User Delay Models from Previous: User Delay Models from
Lin Zhong 2003-12-20