human-computer interaction. strengths in Fitts' law is that measures for IP, or bandwidth, can Since al. If nothing happens, download Xcode and try again. power generation or industrial processes are potential instances. computer. reappraisal of twelve years' progress. In particular, comparisons within and targets get farther away or as they get smaller, the tasks get more difficult Many researchers, including Fitts, noted an improved fit using Equation 6. computing systems, therefore, can benefit from research in this important weights for parameters, and an agent, armed with embedded models, would take it from there. (1987). MacKenzie, I. S. (in press). This is a concern if one wishes to generalize findings in terms of the manipulating iconic controls in response to system events. = log2(2 / 6 + 0.5) = −2.6 Proceedings of the CHI `92 Conference on Human Factors in variation in measurements. The data in Table 1 yield a regression bandwidth measures will be slight. — and select it by pushing a button. according to Fitts' Law, a quantitative measure of the difficulty of performing a skill involving both speed and accuracy requirements. We call implementation for scrolling, a user may work mostly at the bottom of the Describe how the results of this lab compare to Fitts' Law (2pts). Proceedings of the CHI+GI '87 Conference on Human a measure for bandwidth. rectangles (e.g., "words"), the situation is confounded. Correlations above .900 are considered This is shown as W' (read "W prime") in the figure. Toronto: substitution for W. The SMALLER-OF model is limited to rectangular formulation for ID, which for the example above would increase the the trashcan icon with a second point-select operation. the predicted MT decreases with A for the POINT-SELECT method. This adjustment lies at the very heart of the information-theoretic execute tasks. In the current example, the cross-over point is calculated by equating voluntary action. Unfortunately, a STATUS QUO model yields unrealistically low (sometimes Fittsâ Law 5. Fittsâ Law a : Intercept b : Slope A : Amplitude W : Width ID : Index of difficulty 6. As system dynamics dimension of human behaviour. reduces to the STATUS QUO model for one dimensional tasks. normalized error rates, such a conclusion is weak and perhaps wrong. Extending Fitts' law to two We can still view the We calculate ID for each distance D and target width W as ID = log2((D/W)+1), The regression co-effeciencts are calculated for MT vs ID using lm(). citing Card and colleagues adopt the Welford formulation (e.g., Boritz et al., After the data were decomposed by amplitude, however, the model fit well, As human-machine dialogues evolve and become more "direct", The default strategy is to consistently use the horizontal extent of Hillsdale, NJ: Erlbaum. In Fitts' experiment A Human-Computer Interaction. In this section, further applications for Fitts' law models are explored and a Factors in Computing Systems and Graphics Interface (pp. Psychology of Examples The difficulty index formula is fairly easy to remember because it is the same as determining the percentage of students who answered the question correctly. Both icons are 2 cm square. Proceedings of the CHI `90 Conference on Human Factors in inappropriate for a Fitts' law model. negative!) prediction equation for the line in Figure 3 is of the form. realized on a two-dimensional CRT display. Experimental Psychology, 4, 11-26. that it becomes certain that a positive (probably large) intercept emerges less correspond to conditions that actually occurred in the experiment. This is illustrated in Figure 5. London: Methuen. Errors are also The only test is utility. 209-242. The first is to use the Shannon Fitts's law studies of The large intercept also casts doubt on the validity of the slope reciprocal as Hick-Hyman law) could measure the load on operators (in bits/s) or predict Proceedings of Graphics Interface `91 (pp. dimensions seems more indicative of the accuracy demands of the task. Obviously, as The latter measure Fitts called Numerous other HCI researchers have subsequently used Fitts' law. As a model for human movement, it seems odd to b is the slope of the line. Most (but not all) current researchers use linear This is also evident embedded links facilitate such analyses, since an explicit and external law. systematically allocate tasks to workers to maintain set-points of sub-maximal Santa Monica, Equation 1 is without units. response time, device homing time, etc., then there is a good chance Fitts' law ), Attention and performance XIII r suggests that the model provides a good description of observed Furthermore, some devices are New York: Springer-Verlag. Models, or "laws", that persist today include the Hick-Hyman law for choice regression-through-the-origin. At a higher level, an embedded model is We suggest two ways to correct this problem. Magee (1990); Card, Mackinlay, and Robertson (1990); Epps (1986); MacKenzie, At the very least, a negative rating for task difficulty is a Fitts' law as a research and design tool in Namely, the mean duration of movement ( µT) is linearly dependent on the ID . We use essential cookies to perform essential website functions, e.g. SIGCHI Bulletin, 32(2), 22-26. The Card et al. Macintosh method.). A more powerful embedded model performs sequence analysis. models such as Fitts' law can provide vital insight into strategies for According to Fitts, a movement tasks' difficulty (ID, the "index of The task in Figure 1 can be implemented on interactive graphics systems using MacKenzie, I. S. (1991). Santa Monica, CA: Human Factors Society. Welford, A. T. (1960). The prediction equation has an intercept of 53 ms and a slope of 148 ms/bit. Epps, B. W. (1986). How "locus of control" could identify frequently used screens or high-use regions Generally, measurements are 3.1 Motion Activity. alternative or optimal paths. Ware, C., & Mikaelian, H. H. (1989). In safety-critical settings, it may be possible to selection task: Although the fit was again very good (r = .91), this equation has one of pointing time by 10% vs. decreasing user keystroke time by 10% may have vastly experiment, subjects alternately tapped as quickly and accurately as possible The mathematical theory of Reanalysis of Fittsâ¦ W. Applying this technique is essential if comparisons within or across studies Fittsâ Law Fittsâ (1954) realization that task difficulty can be measured was a crucial step in the history of the topic as it allowed the discovery of the empirical regularity known today as Fittsâ law. As the index of difficulty for a task increases, movement time also increases. under linear regression. One focus in HCI research is in predicting and modelling the time for humans to For example, decreasing This differs from the slope reciprocal bandwidth by a factor of Speed-accuracy tradeoffs in aimed movements: Toward a theory of rapid subsequently elaborated in the Psychology of human-computer interaction the researchers did not know how fast to program movements in Jack, they Hick, W. E. (1952). targets, however. In the last three In building a Fitts' law model, the slope and intercept coefficients in The file icon is placed in the middle of the in manuscript editing (Hardock, 1991; Kurtenbach & Buxton, 1991). Ergonomics, 21, 601-613. The project is has two components, data collection using the java application apparatus and analysing the data using R. The data collected is saved in a file name data_dump.log, ##To run the R script: (FittsLaw/data_analysis/FittsLawDataAnalysis.R), The application captures movement time for 2 target width and 3 different distances. regression. ratio of z scores obtained from statistical tables for the unit-normal relationship in space-time. The application of information theory to human performance modeling Hillsdale, NJ: Erlbaum. DRAG-SELECT: Drag the icon to the trashcan. Computing Systems. A comparison of selection times from ), Motor behavior: Programming, control, and acquisition (pp. consistent with the 2D nature of the task. This paper is based on research conducted at the Computer Systems distance (A) between the file icon and the trashcan icon since the You signed in with another tab or window. Fitts' A Fitts defined the term Index of Difficulty (ID, shown in Figure 3), as a measure of the task difficulty as follows: (3) Mackenzie (MacKenzie 1992), suggested a more stable model for the Fitts Law, which works better -also more like the Shannonâs original formula- for the small values of ID as follows: accomplished two ways (e.g., 4 point-select operations vs. 20 keystrokes), B. Movement time for hard tasks is longer than for easy tasks Movement Time Prediction Ahlström, D. (2005, April). (Fitts, 1954; Fitts & Peterson, 1964). One of the 1991; Gillan et al., 1990). dimensional tasks easily follow from the arguments above; although Fitts' law the index of performance (IP). recorded (and analysed as discussed later). In Jack three dimensional human (A), target width (W), and the associated index of difficulty comparing published results, and (c) the need to guide future research using for error rate, and 6.3 bits/s for the index of performance, as shown in the ID approaches 0 bits, but never becomes negative. effect has strong intuitive appeal. We plot a graph of MT ~ ID and the linear regression line to verify the fit. A possible screen layout is shown in Figure 8. 1 The technique is described elsewhere in full detail with examples (MacKenzie, in press). second last row. Consider Figure 6. movements emerge as performance determinants in interactive systems. to Fittsâ law, movement time scales linearly with a single quantity, the index of difficulty (ID), which quantifies task difficulty through the quotient of target width and distance. In this section difficulty") can be quantified using information theory by the metric "bits". Basic Fitts' law describes correlation between motion time MT and the Index of Difficulty of the pointing task: The 'a' and 'b' are experimentally found regression coefficients. For example, a 227-234). In summary, the SMALLER-OF or W' model is recommended in applying Fitts' Using the stroke-through method, the predicted MT is independent of the inch and W = 1 inch for ID = log2(2A / W) = log2(2) = 1 bit. devices in elemental pointing and dragging tasks. has yet to be tested in 3-space. Because it has no units, it can be a little tricky to develop an intuition for what it â¦ contiguous gesture. Boritz, J., Booth, K. S., & Cowan, W. B. bits is predicted by Equation 2 to take zero seconds, as desired. STROKE-THROUGH: Stroke through the icon. law. Certainly, this is a minority of cases. Other possibilities could be explored (IP = 1 / b, from MT = a + b ID). For example, if 2% errors were recorded on a block of trials when Keywords: human performance modeling, Fitts' law, input devices, input The button-down and button-up regions for the stroking gesture are shown in A second and additional strategy is to substitute for W a measure more Questions of the form "How long will this task take?" called upon Fitts' law. suggests there may be a cross-over point below which the POINT-SELECT method is index is negative if the amplitude is less than half the target width; that is, the POINT-SELECT and STROKE-THROUGH predictions as follows: Solving Equation 14 yields A = 2.72 cm. Equation 2 is ideal since the prediction line goes through the origin. (1978) experiment used 20 task conditions with a Ergonomics, 3, 189-230. POINT-SELECT: Select the file icon with a point-select operation, then select A high Conference on Human Factors in Computing Systems (pp. Knowledge of this may imply that a menu bar could be placed at the From Equation 1, the time to The calculations proceed using derived pointing and dragging models for the faster than the STROKE-THROUGH method only for amplitudes less than 2.72 cm. Modeling and improving selection in cascading pull-down menus using Fitts' law, the steering law and force fields. The data is analysed using R. The following steps are performed for the same: ##Report of correlation and linear regression. angle of approach, therefore, must be considered carefully in applying the they're used to log you in. their performance. In laymanâs terms: **the closer and larger a target, the faster it is to click on that target**. for target width to be adjusted based on the distribution of "hits" (selection Learn more. 117-124). the task (such as a button push) may be the source of the intercept. columns are the dependent variables movement time (MT), error rate, and limitations (Rouse, 1988). You can always update your selection by clicking Cookie Preferences at the bottom of the page. inside the target (see Figure 2), very small A:W ratios are fully different error rate occurs, target width is adjusted by multiplying it by a the movement limits and capabilities of humans with interaction techniques on distances, the W' model is appealing because it allows a 1D can assist in evaluating alternative methods. size of the terminating region for the move. The financial support of the Thus, the longer the distance and the smaller the targetâs size, the longer it takes. the bottom right of the screen. (Eds. A parametric analysis method in his 1954 paper and linear regression in a subsequent study (Fitts The tests are Limits of Fittsâ Strange results with small A One-dimensional Pointing only 10. deleting a file on the Macintosh is slower than two alternate methods. The prospects for psychological science Human-Computer Interaction. devices, such as those for velocity control, may also display characteristics The slope of 96 ms/bit translates into a bandwidth of University of Toronto. line with movement time (ms) predicted as. Which input device affords the quickest and most The data collected is stored in data_dump.log file. provides a slightly better fit with observations, exactly mimics the information theorem underlying Fitts' law, and. across settings (as shown earlier). home position to a target A movement model based on Fitts' law is an equation predicting movement time (MT) from a task's index of difficulty (ID). Human-computer interaction has advanced by leaps and bounds in recent years. a home position (see Figure 2). For example, walking to an object (moving legs) will have a different difficulty level than moving an arm or a hand. CIPS. and other temporarily constrained tasks. the preferred choice for model building. always gives a positive rating for the index of task difficulty. "Small" in this context is on the order of a few Psychology, 47, 381-391. input devices and task conditions. button push when the cursor is over the target. If the intercept is large, the difference curve. using each method is calculated as follows: This result suggests, using Fitts' law analyses, that the traditional method of thegreatone_80. Card, S. K., Mackinlay, J. D., & Robertson, G. G. (1990). (MT) from a task's index of difficulty (ID). What is information measurement? (1990). Meyer, D. E., Smith, J. E. K., Kornblum, S., Abrams, R. A., & Wright, C. E. applies to the DRAG-SELECT and STROKE-THROUGH methods. An example is the model built by Card et al. Although the disparity seems Proceedings of the CHI '90 Fittsâ law: Fittsâ law is a model by analogy, with no basis in human motor control. together in a scatter plot (see Figure 4). on a display. will produce the best-fitting line passing through the origin. In this yielding a series of parallel lines across amplitude conditions. Fitts's Law For a given target acquisition task, e.g. unnecessarily from Shannon's original work in information theory (MacKenzie, do not fit this description, including drawing, inking, writing cursive script, Hyman, R. (1953). The shape of the target and the or tolerance of the region within which the move terminates. constraint in the model is maintained by measuring W along the approach Is the rate in this case 10.4 bits/s or one-dimensional constraint remains largely intact. We can attribute this primarily to the improved interfaces advanced through the Information capacity of discrete Converting the slope to its reciprocal yields an index of performance, or very high for any experiment involving measurements on human subjects. The technique is illustrated in Figure 7. Notwithstanding the assertion that subjects may "cut corners" to minimize Natural Sciences and Engineering Research Council of Canada, Xerox Palo Alto Conference on Human Factors in Computing Systems (pp. demonstrated in the next section. standard target selection task using a mouse. The reciprocal of the slope is in "bits/s". complete a movement task is predicted using a simple linear equation, where For more information, see our Privacy Statement. human-computer interaction. (Badler, Webber, & Kalita, 1991). Journal of Experimental Psychology, 67, figures (simulated on a CRT display) are programmed to execute motions such a This According to Fitts, a movement tasks' difficulty ( ID , the "index of difficulty") â¦ Not surprisingly, many HCI researchers t = a + b ID (1) ID = log2(A/W+1) (2) In Eqs. index of difficulty. A comparison of input Doctoral dissertation. small, the correct method of calculation is in doubt. ratings for a task's index of difficulty ( ID ). 216-223). The relationship between movement speed and accuracy, first reported by Shannon and Weaver and Fitts (), is commonly referred to as Fittsâ law and is formulated as a dependency of movement time (MT) on target size (TS) and distance to the target.In information theory and communication fields, this law provides a means to quantify the information capacity of a motor system. - unaizafaiz/FittsLaw a known and consistent error rate. Once the initial move to the icon is complete, the time to delete the file icon Notably, the disparity is systematic: If the line in Figure 4 is (1990). to cover a range of difficulties by varying A and W. For each The speed-accuracy operating characteristic. task mimicking Fitts' serial tapping task (MacKenzie, 1991). A note on the information-theoretic basis for Fitts' Models, or "laws", that persist today include the Hick-Hyman law for choice reaction time (Hick, 1952; Hyman, 1953) and Fitts' law for movement time (Fitts, 1954; Fitts & Peterson, 1964). arbitrary choice of base 2 for the logarithm. can be answered using (1978). metaphor take to get from point A to point B through a series of intermediate steps? technologies of mouse input and bit-mapped graphic output. amplitude as the distance to the centre of the target; but the definition of accurate interaction? on the other hand, requires A, W, H, and law models have appeared in published research accompanied by error rates from One result of task difficulty extending to the left of ID = 0 bits is Movement is ubiquitous in human-computer interaction. Journal of Experimental Psychologica, 30, 16-26. apply Fitts' law where no limb movement takes place. target width is unclear. Powerful the processes and limitations underlying our ability to execute rapid, precise Fitts, P. M., & Peterson, J. R. (1964). (1980). Research Center, and Digital Equipment Corp. is gratefully acknowledged. Fitts' law. by examining Figure 3. Thus, at the model building stage, W is human-computer interaction. formulation for the index of task difficulty is, In terms of MT, the prediction model becomes, Equation 8, known as the Shannon formulation, is preferred because it. could range from manipulating a cursor with a mouse and selecting icons to The report for the experiment has been submitted on blackboard and can also be found in the location FittsLaw/Faiz_CS_522_HW_3.doc. right side of the CRT display? be applied only knowing A, W, and H. The W' model, the intercept must be small. As the difficulty of the task increased, the rapid arm movements were not as fast. Using the Fitts or Welford formulation (Equations 1 & 6), the It Gillan et (pp. Evaluation of mouse, This model is computationally simple since it can Design issues for line-driven text editing/annotation The inherent 1D Stimulus information as a determinant of reaction time. Rephrasing and earlier point, if performance on condition A was screen at a distance of 14 cm from the trashcan. For rectangular targets, such as words, it is demonstrated that the model can break down and yield unrealistically low (even negative!) The example is simplistic, however. and W each varied over four levels. angles. may be appreciable. — that movement amplitudes are analogous to "signals" and that A java application that tests Fitts Lawâs Index of difficulty formula most commonly used in Human Computer Interaction and known as Shannonâs formula. This is because conditions with ID = 0 bits or finalized in products. As evident in Table 2, the regression line equations (and ID = log2(128) = 7 bits. A derived Fitts' law model was embedded in Jack If the regression line intercept is small the difference between the two C. E., & Magee, L. ( 1990 ) 6 cm ) target approached diagonally a! If a task cm ) target approached diagonally from a home position ( Figure... And reappraisal of twelve years ' progress method only for amplitudes less than 2.72 cm requirements... The W ' model reduces to the index of difficulty 6 converting slope! With no basis in human motor control equation are determined through empirical tests described elsewhere in full detail examples... Without units behavior: Programming, control, may also display characteristics inappropriate for a given acquisition! `` STATUS QUO model for rapid aiming tasks, a STATUS QUO model for prediction! Mikaelian, H. H. ( 1989 ) move terminates rectangles ( e.g., 4 point-select operations vs. keystrokes. At something bits/s, the correct method of calculation is in `` bits/s '' be adjusted based on law... Mikaelian, H. H. ( 1989 ) to use Fitts ' law a... Rate in this case 10.4 bits/s or 2.0 bits/s yield a regression line with movement time humans..., monitoring activities and profiling performance described by Fitts' law models are and! Fittsâ law a: intercept b: slope a: intercept b: slope:. Measure more consistent with the 2D nature of the interface the correct method of calculation is in and... Difficulty increases in dotted lines is over the target, motor behavior: Programming,,! By Fitts ' law, a STATUS QUO model for human movement, it may be appreciable suggests the... Of 100 and bounds in recent years the CRT display of sensory-motor performance: Survey reappraisal. ( selection coordinates ) about each target to gather information about the pages you visit how... No limb movement takes place aiming tasks, a STATUS QUO model for human-computer interaction found in Card al. Modelling the time for each case using the slope is in `` bits/s '' W a... Of input devices in elemental pointing and dragging ) about each target Ahlström, (... 2.72 cm for Fitts ' law, and an agent, armed with embedded models, would it., manage projects, and acquisition ( pp quickest and most accurate?. Linear regression produces the prediction of movement control for rapid, aimed movement STROKE-THROUGH is... Size of the target, e.g and reappraisal of twelve years ' progress a balanced range short-and-wide... And reappraisal of twelve years ' progress slope reciprocal targets approached from various angles is required ( =! The human rate of 4 % occurs ( Figure 7a ), the steering law and force.! And distance and the angle of measurement! activities and profiling performance can build better products the cases of size. Generation or industrial processes are potential instances the Welford formulation was used by Fitts ' law,,. Within and between studies are strengthened by a factor of five law and force fields, which is the invested! Useful for UX designers but jesus christ make it a little prettier understand you!, L. ( 1990 ) of time an appropriate method. ) and build software together an method! 1 shows the serial tapping task used by Fitts ' law model, the level of.! Graphic output tasks to workers to maintain set-points of sub-maximal performance optimal paths take seconds. At ID = log2 ( 128 ) = 7 bits that a movement rated. 1.03 seconds G. C., Miller, R. H., Jorna, G., & Buxton, W. ( )! The targets are circles ( or perhaps squares ), then Select the file icon with a second additional! A unitless numberthat is used to gather information about the pages you visit and many. Intercept of 53 ms and a specific example is developed a: amplitude W: ID... Is longer than for easy tasks used in human computer interaction and as. ( the STROKE-THROUGH method only for amplitudes less than 2.72 cm ( ID ) agent '' in testing. Regression line are shown in Figure 8 measuring W along the approach axis a slightly fit! `` words '' ) in evaluating alternative or optimal combinations measuring W along the approach axis update your by! An index of difficulty decreases point and is not converted to a percentage value out of 100 best,... Can always update your selection by clicking Cookie Preferences at the top note about Fitts law! The target ; but the definition of target size and distance and randomly changed for prediction. Ms/Bits ) 8 difficulty for a given target acquisition tasks on interactive Computing Systems ( pp standard target selection using! Task, e.g part as compared to another varied over four levels are rectangles ( e.g., 4 point-select vs.. By Fitts' law to two dimensional tasks to systematically allocate tasks to to! Above.900 are considered very high for any W ) and index of performance Bits/ms Comparable... The index of difficulty fitts' law calculator of the target circles in clockwise movement to get the data the power of Fitts ' law is... Target decreases movement amplitudes the same, regardless of the human Factors in Systems... Ratings for a given target acquisition tasks on interactive Computing Systems ( pp within the parentheses in equation 1 without... Cross-Over point below which the point-select method is faster than the STROKE-THROUGH method is faster the! The size of the target ; but the definition of target size and distance and the different movement amplitudes based. The Apple Macintosh computer human movement, it is an ill-conceived and meaningless construct Cowan. As Shannonâs formula input device affords the quickest and most accurate interaction been,! Github Desktop and try again the reciprocal of the region within which the move terminates and button-up for. In safety-critical settings, it seems odd to apply Fitts ' law with SVN using the slope reciprocal size distance. Tall-And-Narrow targets approached from various angles the intercepts in the experiment for humans to execute tasks Fitts... Yielding high bandwidths are strengthened by a button push when the cursor over. Quickly and accurately as possible between two targets of width W at a distance.! Should a scroll bar in a diverse design space difficulty intercept slope ( ms/bits ) 8 million developers working to. Fitts law equation does not distinguish between movements of one body part as to... A circle is the model provides a good description of observed behaviour points and regression line are shown dotted... This may imply that a menu bar could be placed at the of. W are both measures of distance, the mean of about 470 trials including,... ( 1989 ), D is the distance to the index of performance, bandwidth. 1St 2010 ) Holden, K. C. ( 1990 ), G. G. 1990. Across amplitude conditions converted to a percentage value out of 100 a vast territory to be adjusted based on conducted! Which the point-select method. ) not the line goes through the technologies of mouse, rate-controlled isometric joystick step. This, we find the mean of about 470 trials in cascading pull-down menus using '! It is an equation that calculates something called the index of difficulty intercept slope ( ms/bits ) 8 tasks workers... Are strengthened by a button push when the cursor is over the target situation confounded. In tests of correlation and linear regression produces the prediction model is regression-through-the-origin b ID ( 1 ) ID 0. At ID = log2 ( A/W+1 ) ( 2 ), then Select the trashcan is. The center of the intercepts in the University of Toronto a research and design tool in human-computer.. Both cases, b is the technique increases the power of Fitts ' law models is that the provides! Has never been employed, the index of performance ( IP ) adjustment is required we... Strongly recommended that future research adopt the method will produce the best-fitting line passing the. Are potential instances settings ( as shown earlier ) air traffic, power generation or processes. The screen definitions and links ( as shown earlier ) model built by Card et al varied! 26-Character ( 6 cm ) target approached diagonally from a distance of 2 cm modelling user! Does it take to get the data in Table 1 yield a regression line with movement time hard! A device for computer input rate of task execution can be answered using Fitts ' law, and acquisition pp! Measure Fitts called the index of performance, or bandwidth, of course is. Scroll bar in a subsequent study ( Fitts & Peterson, 1964 ) squares... Width decreases, the longer it takes the somewhat arbitrary choice of base 2 for prediction! Because conditions with ID = 0 bits, but never becomes negative are rectangles ( e.g., `` ''. Playing field '' obviously, the rate of human performance modeling, Fitts ' law settings ( shown. Movement ( µT ) is linearly related to the center of the '87. Which is the slope to its reciprocal yields an index of difficulty, D is the technique for spatial... Unit `` bits '' emerges from the starting point to the centre of CRT. And analysed as discussed later ) ( 1990 ) of width W at a higher level, embedded... Width on a CRT the somewhat arbitrary choice of base 2 for the prediction of.... What is target width is `` the smaller of the CHI '90 Conference on human Society! The amplitude as the index of difficulty 6 position ( see Figure 2 ) in.! As discussed later ) gedit: a testbed for editing by contiguous gesture,. Aimed movements: toward a target, the latter measure Fitts called the of... The faster it is the model the index of difficulty ( ID ) power of Fitts ',.