"... In PAGE 5: ... We measured spatial memory by the number of objects correctly recalled after completion of the memory task as our dependent variable; our independent variable was the type of interaction as before. Table2 shows the descriptive statistics for our measurements. Comparing the means of the spatial memory scores in the two types of interactions using one-way ANOVA, we found that the number of correctly recalled objects in eye movement- based interactions was indeed significantly lower than that in pointing (F[1,46] = 7.... ..."

"... In PAGE 5: ... We measured spatial memory by the number of objects correctly recalled after completion of the memory task as our dependent variable; our independent variable was the type of interaction as before. Table2 shows the descriptive statistics for our measurements. Comparing the means of the spatial memory scores in the two types of interactions using one-way ANOVA, we found that the number of correctly recalled objects in eye movement- based interactions was indeed significantly lower than that in pointing (F[1,46] = 7.... ..."

"... In PAGE 3: ... The system is initially in scanning mode, which requires a set of events to occur to switch into reading mode. The events that are tracked include the specific eye movements shown in Table1 . For example, if the eye moves a short distance left then the event is a regression saccade but if the eye moves a long distance left then the event is a scan jump .... In PAGE 3: ... For example, if the eye moves a short distance left then the event is a regression saccade but if the eye moves a long distance left then the event is a scan jump . The distinction among short , medium and long distances used to characterize events (in Table1 ) reflect a set of adjustable parameters, one for each distance and each direction pair. The quantized, tokenized stream of eye-movement data is then pooled to determine whether the user is reading.... In PAGE 5: ....1.2 Results The raw gaze data were analyzed with both our pooled evidence algorithm and the simple algorithm based on Jacob apos;s fixation recognition method. For the pooled evidence algorithm, data were first quantized and then tokenized according to the rules in Table1 . The number of tokens was counted for each condition and subject.... In PAGE 7: ... A second method to enhancing our algorithm would be to actively adapt its parameters. We will include parameters that adapt to individual reading speeds and abilities by adjusting parameters that are used to determine the actual values of the distances we called short , medium , and long in Table1 . If, for example, the system determines that the user is a slow and careful reader, all the distances (for the x axis) should be decreased to optimize performance.... ..."

"... In PAGE 3: ... The system is initially in scanning mode, which requires a set of events to occur to switch into reading mode. The events that are tracked include the specific eye movements shown in Table1 . For example, if the eye moves a short distance left then the event is a regression saccade but if the eye moves a long distance left then the event is a scan jump .... In PAGE 3: ... For example, if the eye moves a short distance left then the event is a regression saccade but if the eye moves a long distance left then the event is a scan jump . The distinction among short , medium and long distances used to characterize events (in Table1 ) reflect a set of adjustable parameters, one for each distance and each direction pair. The quantized, tokenized stream of eye-movement data is then pooled to determine whether the user is reading.... In PAGE 5: ....1.2 Results The raw gaze data were analyzed with both our pooled evidence algorithm and the simple algorithm based on Jacob apos;s fixation recognition method. For the pooled evidence algorithm, data were first quantized and then tokenized according to the rules in Table1 . The number of tokens was counted for each condition and subject.... In PAGE 7: ... A second method to enhancing our algorithm would be to actively adapt its parameters. We will include parameters that adapt to individual reading speeds and abilities by adjusting parameters that are used to determine the actual values of the distances we called short , medium , and long in Table1 . If, for example, the system determines that the user is a slow and careful reader, all the distances (for the x axis) should be decreased to optimize performance.... ..."

"... In PAGE 4: ... Our dependent variable was performance (time to complete task). Table1 provides the descriptive statistics for our measurements. We tested the first hypothesis by comparing the means of the pooled performance scores (performance scores in close and distant virtual environments) using one- way analysis of variance (ANOVA).... ..."