The times for relatively experienced users entering the five formulae into LATEX, Microsoft's Equation Editor (MSEE) and the Freehand Formula Entry System (FFES) are presented in the following table. Times are measured in seconds. For comparison, times for the user testing's participants are shown as well, indicative of a novice user's speed.
The times shown for the system (FFES) is the raw entry plus the time taken for corrections of the occasional grouping and recognition errors. Because of the poor performance of the formula parser, the time for parsing the formulae is not included. It would unfairly penalise thesystem due to the fact that the formula parser could be replaced with one that was faster. These times are intended to indicate the entry times for formulae.
It is important to note that the novice users of the system were not attempting to achieve fast entry times. This data reflects the time of unhurried formula entry, and is included here as a rough upper bound. The ``Best Novice'' times are the individual best novice times for each formula, selected from all the eight participants in the user testing.
For equations that were near-linear in structure, entering straight LATEX or using a template-style equation editor, such as Microsoft's Equation Editor , proved to be faster than using the system. For more complex equations that needed to be ``laid out'' in 2D, entry time for a user of this system was slower, yet comparable, to that of a relatively experienced user of more conventional systems. In comparison to Microsoft's Equation Editor, users of the system found the entry of formulae to be easier and much less frustrating, particularly if they were doing editing operations.
The novices' average times are much higher than those of the experts, primarily due to low character recognition, averaging 77%. Their unfamiliarity and tentativeness with the pen and tablet interface, as well as not having trained the character recogniser, are the main reasons for this. The two novice users who had prior experience with a pen and tablet performed much better than the averages suggest.
, proved to be faster than using the system. For more complex equations that needed to be ``laid out'' in 2D, entry time for a user of this system was slower, yet comparable, to that of a relatively experienced user of more conventional systems. In comparison to Microsoft's Equation Editor, users of the system found the entry of formulae to be easier and much less frustrating, particularly if they were doing editing operations.
The novices' average times are much higher than those of the experts, primarily due to low character recognition, averaging 77%. Their unfamiliarity and tentativeness with the pen and tablet interface, as well as not having trained the character recogniser, are the main reasons for this. The two novice users who had prior experience with a pen and tablet performed much better than the averages suggest.