The introduction of the computer into the office environment has resulted in drastic changes in the way business is conducted. The computer system's display device has the highest bandwidth for the delivery of data to the user, and at the present time, some, if not all of that data is transferred through the medium of text. This study will focus on a human performance measure of the effectiveness of display devices in the presentation of text.
Rupert (1991) notes that "while it is very tempting to strive for ever increasing bandwidth to the office, a logical upper bound is imposed by the maximum information which a individual can process. Thus a basic criteria to insure that future needs will be satisfied is that we match the maximum bandwidth of a new system to the bandwidth which an individual can perceive." He estimates the capacity of the human optical system at between 1 and 10 GBits/sec. This far exceeds the audio input capacity of approximately 0.6 Mbits/sec. In an age of increasing capacity in system processing speed and memory, the display on which data is to be presented represents a bottleneck for system designers. Silzars (1995) has estimated the resolution necessary for a display to achieve the same visual quality as a 35 mm still photograph - between two to four megapixels. With advanced display technologies, these are achievable, although quite expensive. Holzel (1997) reports that the current state-of-the-art in advanced flat-panel displays provides excellent rendition of still images, but falls short for display of moving images because of shortcomings in one or more of the following areas:
Defining the effectiveness of a system in terms of human performance requires the definition of a standard task which can provide reproducible results. Computer systems today are used to provide information, services, socialization, and entertainment. We propose measuring reading speed as a means of defining the system's potential in terms of human performance. With the overall bandwidth restrictions on today's "information superhighway", it is likely that, for information seeking, text will remain a significant mode of information delivery.
Even though the process of reading text is extremely complex, the word "reading" has become synonymous with "comprehension", or getting meaning out of text. Myer (1981) identifies nine variables affecting prose comprehension - "the content of the information in a text..., the structure of this content...emphasis (which) reflects an author's purpose, attitude, or mood...inference, which authors assume their readers can make to fully comprehend their text...the learner's world knowledge...the perspective of the learner...the learner's purpose...the skill level of the learner...and finally, the predominant processing style of the individual learner..." Myer (1981) . Some of these variables are properties of the textual material itself, some are properties of the situation, and some are properties of the reader. Because of this broad range of variables, measurements of reading performance often fail to provide a quantitative, reproducible measurement of what is learned in the process of reading.
Marcel Just (1996) has developed a modeling system .(3CAPS) to provide a unified formalism to express the dynamics of high-level cognitive processing in sentence comprehension. A distinguishing feature in the model is its limited resource pool for activation, which provides the ability to model individual differences in the speed and correctness of thinking. The test we propose for this study will limit the resource utilization of the subjects, and should minimize the higher level cognitive processing that lead to the wide variations in performance noted by Myer. LaBerge and Samuels (1974) have termed the process of converting the lines that represent characters into the characters themselves "automaticity", and hypothesize that, while beginning readers may spend up to 90% of their brain's available processing time on this process, more experienced readers will spend less than 10% of their time in this process, freeing mental resources to deal with the conversion of letters into words, and the processing of those words into concepts or information (comprehension).
If this test allows system designers to determine the benefits of the choice of a particular type of display in terms of human performance, separate from the performance enhancements under software control, and, therefore under the control of the end user, the test, and the process of collecting this data should be useful.
Continue
Return to the Title Page for Comparison of
AMLCD and CRT Displays