CAR-TR-506
`CS-TR-2487
`
`June 1990
`
`A new era for touchscreen applications:
`High precision, dragging icons, and refined feedback
`
`Andrew Sears, Catherine Plaisant, Ben Shneidennan
`
`Human-Computer Interaction Laboratory
`Deparnnent of Computer Science
`University of Maryland
`College Park, MD 20742
`
`(To appear in: Advances in Human-Computer Interaction, Vol. 3, R.Hartson, D. Hix, Ed.)
`
`SCEA Ex. 1026 Page 1
`
`

`
`A new era for touchscreen applications
`
`- 1 -
`
`A new era for touchscreen applications:
`High precision, dragging icons, and refined feedback
`
`Andrew Sears, Catherine Plaisant, Ben Shneidennan
`Human-Computer Interaction Laboratory
`Department of Computer Science
`University of Maryland
`College Park, Maryland
`
`June 1990
`
`(all the figures are at the end of the document)
`
`1. Introduction
`One goal of human-computer interaction research is to reduce the demands on users when
`using the computer. This can be done by reducing the perceptual and cognitive resources required
`to understand the interface or by reducing the motor effort to use the interface. The introduction of
`alternative input devices, such as the mouse and joystick, significantly improved some user
`interfaces. The touchscreen combines the advantages of these other devices with a very direct
`method of inputting information. Users simply point at the item or action of interest, and it is
`selected.
`While many input devices allow interfaces to be customized, increased directness distinguishes
`touchscreens. Touchscreens are easy to learn to use, fast, and result in low error rates when
`interfaces are designed carefully. Many actions which are difficult with a mouse, joystick, or
`keyboard are simple when using a touchscreen. Making rapid selections at widely separated
`locations on the screen, signing your name, dragging the hands of a clock in a circular motion are
`all simple when using a touchscreen, but may be awkward using other devices. Even when a task
`can be accomplished with other input devices, users may have to clear their workspace for the
`mouse or press many keys to move the cursor.
`Touchscreens have long been thought of as being simple to use. Unfortunately they have a
`reputation as being practical only for selecting large targets and as being error prone. Recent
`empirical research, as well as advances in touchscreen hardware, have dramatically improved the
`performance of touchscreens and the range of applications for which they can be advantageously
`used. Even with these advances, today most touchscreen applications emphasize the metaphor of
`
`SCEA Ex. 1026 Page 2
`
`

`
`A new era for touchscreen applications
`
`-2-
`
`'buttons' being pressed on the screen. Tasks such as dragging an object on the screen, moving the
`marker on a slider, or free hand drawing are rarely attempted with touchscreens, but we believe
`that touchscreens can excel in such cases. This chapter presents recent empirical research which
`can provide a basis for theories of touchscreen usage. We believe recent improvements warrant
`increased use of touchscreens. Human factors specialists, psychologists, and computer scientists
`have a grand opportunity to influence further developments and refine theories in these new
`domains.
`
`2. Advantages and perceived disadvantages of touchscreens
`There are many advantages to touchscreens which have made them popular for public access
`situations.
`
`2.1 Advantages
`Directness:
`One of the biggest benefits of a touchscreen is its directness. Unlike indirect devices such as a
`mouse, joystick, or keyboard, touchscreen users simply point at the desired object, and it is
`selected. There is no need to remember a complex syntax, search for the input device, remove
`visual focus from the objects of interest, or press multiple keys to move the cursor. More
`importantly, there is no need for users to map hand motions to cursor motions, as required by
`many other input devices. Sliding, dragging, and gestural input also benefit from the touchscreen
`directness.
`
`Speed:
`The touchscreen is the fastest selection device for many tasks. Users do not need to reach for
`the input device when it is time to make a selection as they often do with a mouse or lightpen. An
`additional advantage in many situations is the lack of a cursor when users are not touching the
`screen. Users simply touch the desired location rather than touching a cursor and dragging it to the
`desired location.
`
`Ease of learning:
`Touchscreens are easy to learn to use. Once users realize that they must simply touch the
`screen to interact with the computer, they quickly master simple actions such as touching buttons
`or dragging items across the screen. Unlike the mouse or tablet there is no need to learn and
`practice spatial reorientation and hand-eye coordination (Nielsen and Lyngbaek, 1990).
`
`Flexibility:
`Touchscreen interfaces offer flexibility not available with a keyboard. Each interface can be
`
`SCEA Ex. 1026 Page 3
`
`

`
`A new era for touchscreen applications
`
`-3-
`
`!'
`
`customized for each specific task performed. Users can choose which keyboard layout they
`prefer, QWERTY, Alphabetic, or Dvorak, since it is displayed on the screen.
`
`No moving parts:
`The lack of moving parts connibutes to the durability of touchscreens that has made them
`popular for applications such as information kiosks at amusement parks, office buildings, or
`museums. Unlike a mouse or keyboard, only the touchscreen must be accessible to users, making
`loss or damage of hardware less likely. One system, an information kiosk developed for the
`Smithsonian, traveled to museums across the country for two years. These touchscreens were
`heavily used and never failed. However, the video monitors did ultimately fail from abuse during
`shipping.
`
`No additional desk space:
`Touchscreens free desk space for other uses. Many input devices, such as the keyboard and
`mouse, require desk space which may be very limited. A related benefit is that the touchscreen is
`in a fixed location. Unlike the mouse or lightpen there is no need to search for the device which
`may be hidden under papers. If the user is currently working with the computer, the screen must
`be accessible. This is particularly useful for applications requiring only occasional pointing.
`
`2.2 Perceived disadvantages
`There are also some problems that have been associated with touchscreens. Many of these
`problems have been overcome or reduced by improvements in touchscreen technology or design
`strategies that have been developed for touchscreen interfaces.
`
`Low resolution:
`This is one of the biggest misconceptions about touchscreens. Many people have reported on
`the low resolution of touchscreens. Some researchers have claimed that the resolution of a
`touchscreen is limited by the size of users • fingers, and others have claimed that selection of single
`characters would be slow if it was even possible. Recent research has shown that targets as small
`as 0.4x0.6mm could be selected with touchscreens (Sears & Shneiderman, 1990). The same
`research concluded that targets 1.7x2.2mm could be selected as fast with a touchscreen as they
`could with a mouse, with similar error rates.
`
`Ann fatigue:
`This could be one of the most significant problems with touchscreens. Using a touchscreen at
`the angle most monitors are currently mounted can lead to arm fatigue, making them difficult to use
`for extended periods of time. Renewed interested in reducing fatigue appears to have resulted in
`
`··.tt.t
`
`'
`
`: nw a _t w:w .. u .. :a .::.mum •.
`
`: •. a s :.:
`
`ooma
`
`IJi@J
`
`(.Ill .. M
`
`tit
`
`.1/d.i.l.&.OH.CJJ.v .J..c.:::d&h.c:f:; ,,, ii1m
`
`, (
`
`JtJiw.X.t.J.,,.
`
`SCEA Ex. 1026 Page 4
`
`

`
`A new era for touchscreen applications
`
`-4-
`
`y
`
`simple changes to the touchscreen position that will significantly reduce this problem (See Section
`6 for more details).
`
`Parallax:
`When touchscreens were first introduced. the infrared technology was prevalent. Early
`infrared touchscreens had the touch sensing devices mounted above the surface of the monitor.
`When users' fmgers were close enough to the screen, the infrared beams would be broken,
`resulting in a touch. This could occur long before the user meant to touch the screen. Newer
`infrared touchscreens, and all other technologies, sense touches much closer to the monitor
`surface, if not directly on the surface, reducing the problem with parallax. Software strategies
`have also been explored that reduce problems created by residual parallax by correcting for offsets
`created by the parallax and providing feedback to users about their exact position.
`
`Glare and smudges:
`Glare and smudges on the monitor are of concern to many designers. Mounting the monitor at
`a better angle, using lightly ground glass surfaces, and paying careful attention to the lighting near
`the workstation can significantly reduce the glare problem. Smudges are unattractive and can
`obscure the display. Reducing smudges simply requires users to clean the monitor occasionally.
`On the other hand we find that some touchscreens have less problems with accumulating dust than
`standard monitors. In our laboratory environment, we find ourselves cleaning the mouse pad and
`mechanical parts more often than we clean the touchscreens.
`
`Obscuring of the screen:
`The fact that users use their fingers to make a selection by touching the screen implies that the
`users' hand will obscure a part of the screen. Careful design of the interface, placing selectable
`items in locations that will keep the user's hand from obscuring the screen, can significantly reduce
`this problem. When possible, the handedness of users should be considered when designing
`interfaces, or users could be allowed to customize the software for the left or right hand.
`
`Limited tactile feedback:
`Visual and audible feedback should be used to compensate for limited tactile feedback in button
`applications. Tactile feedback is particularly important when performing rapid button presses
`without watching the screen. An example is typing on a touchscreen. When users type on a
`traditional keyboard, the edges of the keys help orient their hands and the motion of the keys
`indicates when they are pressed. These cues are not available with touchscreen keyboards. Visual
`and audible feedback can supplement the physical contact with the screen to help compensate for
`the absence of key motion, but identifying when the edge of the touchscreen key is touched is more
`
`SCEA Ex. 1026 Page 5
`
`

`
`A new era for touchscreen applications
`
`. 5.
`
`difficult. When performing tasks that involve sliding and dragging, the friction between the users'
`finger and the screen provides some tactile feedback. Although this problem is not unique to
`touchscreens, it is an important consideration when designing touchscreen interfaces. Currently
`research is being conducted to improve user performance for 'touch typing' ~th touchscreens
`(Sears, 1990).
`
`Undesired touches:
`When using touchscreens users may rest their hands on the screen for extra support or to
`reduce ann strain, or they may inadvertently touch the screen with another finger. This causes
`touchscreen hardware to loose track of the location users wish to touch. Research with
`touchscreens that recognize multiple touch locations may prove useful in eliminating this problem.
`
`Price:
`Touchscreen prices are getting lower, but are still relatively expensive. Touchscreens range
`from approximately $350 to over one-thousand dollars. This is considerably more than most mice,
`joysticks, or lightpens.
`
`Many of these problems have either been overcome or reduced, and usage is steadily
`increasing. Many of the problems associated with parallax and glare have been overcome by
`advances in touchscreen hardware. Design guidelines can significantly reduce the problems
`associated with obscuring the screen, the lack of tactile feedback, and undesired touches. There is
`renewed human factors research into reducing fatigue that appears promising. The price of
`touchscreens is decreasing as technology improves and touchscreen use increases. It is anticipated
`that when manufacturers start to produce monitors with touchscreens installed at the factory, the
`price of touchscreens should drop significantly.
`We cannot resist mentioning some of the historical prejudices against the touchscreen. Many of
`the pioneer touchscreens did have severe limitations. As a result, many people still picture
`touchscreens as low precision, high error rate input devices. Touchscreens can be reliably used to
`select relatively small targets (approximately 2mm square). Target size is not limited by the size of
`the fmger as several researchers have claimed. Touchscreens do not require a large and intrusive
`frame glued or taped on a monitor as many early versions did. They can be mounted directly onto
`the surface of the monitor and all supplemental hardware can be installed inside the monitor. In
`summary, touchscreens have improved dramatically in recent years, and, as a result, high
`precision, low error rate tasks can now be performed using a touchscreen. Now psychologically
`oriented researchers can explore new strategies and applications to guide practitioners.
`
`SCEA Ex. 1026 Page 6
`
`

`
`A new era for touchscreen applications
`
`-6-
`
`3. Comparison with other input devices
`Touchscreens have been compared empirically to mice, lightpens, keyboards, joysticks, and
`other devices. The majority of human factors studies tested touchscreens against various devices
`for selecting predefmed, stationary targets. Time and error rates were measured, and some studies
`measured user satisfaction. In general, these studies have shown that touchscreens are the fastest
`device for selecting stationary targets (Muratore, 1987; Ostroff & Shneiderman, 1988; Ahlstrom &
`Lenman, 1987; Karat, McDonald & Anderson, 1986). Unfortunately, touchscreens have also been
`shown to be the most error prone input device (Muratore, 1987; Ahlstrom & Lenman, 1987).
`However, much of this research emphasized relatively large targets, and few used alternative
`selection strategies that may improve user performance, making it of limited use for higher
`resolution tasks, such as character selection or graphics input
`A recent study (Sears & Shneiderman, 1989) compared the touchscreen to the mouse for the
`selection of various size targets when using the lift-off selection strategy (the lift-off strategy will
`be described in the following section). This study showed that using the ,alternative selection
`strategies can result in very low error rates for the touchscreen. It also showed that selection of
`very small targets (0.4x0.6mm) is possible with the touchscreen, refuting claims that the size of the
`user's finger determines the minimum target size. This study showed that the selection of targets
`the size of a character, or smaller, is as fast with a touchscreen as with a mouse.
`Other studies have compared various input devices for selection tasks when users must also
`type on a keyboard. One study compared a touchscreen, mouse, and keyboard for a selection task.
`Results indicate that the touchscreen was preferred for tasks that do not involve a typing subtask,
`and that the keyboard was preferred when the typing subtask was included. The touchscreen was
`preferred over the mouse when performing a typing subtask. The touchscreen was also the fastest
`device for cursor positioning in both situations (Karat et al., 1986).
`Unfortunately, there have been few comparisons of the touchscreen to other input devices for
`tasks other than target selection. Tasks such as dragging objects and outlining an object offer new
`opportunities for researchers. Informal observations indicate that tasks involving unconstrained
`movements may be significantly easier with the touchscreen than many other devices. Of course, a
`lightpen or tablet may prove superior for some tasks due to their similarity to writing with a pen.
`Additional studies are needed to understand the range of tasks for which touchscreens can and can
`not be used.
`Overall, touchscreens appear to be the fastest device for selecting relatively large targets. They
`can also be used for selecting smaller targets if the correct selection strategies are used. Error rates
`can also be reduced to a point where they are negligible if the correct strategies are used. Tasks
`such as dragging objects on the screen or marking the border of an irregular region, also appear to
`be very promising with touchscreens. The bibliography contains references to many papers
`dealing with touchscreens with indicators of the subject of the papers.
`
`SCEA Ex. 1026 Page 7
`
`

`
`A new era for touchscreen applications
`
`-7-
`
`4. Designing touchscreen applications
`4.1 A model of user interaction
`We might consider a model of operation that divides touchscreen usage into seven stages.
`These are based on the syntactic/semantic (Shneiderman, 1987) and seven stages (Norman, 1988)
`models. The user
`
`1) formulates a plan of what needs to be done in the task domain,
`
`2) examines the current computer screen to identify all touchable and non-touchable areas that
`represent actions and objects relevant to the task,
`
`3) identifies the desired touchable area by the action or object,
`
`4) reaches out to touch (the syntax is simply a touch) the desired area and receives feedback
`from hand position and from on-screen changes (a cursor, selectable areas inverting, etc.),
`
`5) confirms that the finger is on the desired touchable area and lifts-off to activate,
`
`6) confirms that the desired touchable area has been activated,
`
`7) interprets and evaluates the result of the touch in terms of whether the task domain goal is
`furthered.
`
`This model describes interaction when the lift-off selection strategy is used with a touchscreen.
`Simple modifications can adapt this model for other selection strategies, or for other input devices.
`The central benefit of using a pointing device rather than a keyboard is reducing the syntactic
`load by replacing typing with pointing. The directness of touchscreens further simplifies the task
`by allowing users to simply point directly at the object or action. Instead of detailed instructions
`about what to type (the syntax) to select an action or object, users simply touch a visual
`representation of the object or action (the semantic description). Touchscreens avoid the distraction
`of looking from the screen to the input device and back while remembering the desired syntax. If
`designers choose a proper set of touchable objects and actions, then progress can be rapid with low
`error rates.
`The traditional touchscreen strategy has been Iand-on, in which the first location that users'
`touch initiated an action. This is acceptable, even preferable, when there are a few large targets.
`However, as the task domain complexity increases, the number of choices can increase
`dramatically and undermine the efficacy of the Iand-on strategy.
`With high resolution touchscreens that support continuous feedback, the finger touch may
`produce a cursor that can be dragged across the screen, and activation occurs when the finger is
`
`, t.to: a ;;smw.
`
`h
`
`. iMWWAWI.
`
`tt ... tt.nmt~.m
`
`. ;.t
`
`XiittJ..t.ttta .. hi\
`
`( ,JS.Jiv.QQiw_hJ c
`
`~ ;
`
`,
`
`iiil(
`
`i!ln.ttEJ\i&.o.n .. ' ... th .•.. w.?.G.bi&.M.w.S,C .. J:
`
`SCEA Ex. 1026 Page 8
`
`

`
`A new era for touchscreen applications
`
`-8.
`
`lifted off the swface. This can not only improve the selection of menu items or buttons but also
`open a new world of interaction techniques such as the direct manipulation of metaphors (moving a
`cursor on a slider, selecting a color on a color wheel, etc.), free hand-drawing, and symbolic
`gestural input. There are many other possible strategies and combinations of strategies that can be
`used.
`The following sections will discuss various interaction techniques, and how each technique can
`be used with a touchscreen.
`
`4.2 Traditional button or menu selection
`Menu or button selection tasks typically require the selection of predefined targets represented
`on the screen.. The majority of current applications for touchscreens involve tasks such as these.
`This section discusses factors that play important roles in button selection tasks including the visual
`representation, size, and location of targets and several alternative selection strategies.
`
`• Visual representation of touchable areas
`Users are constantly confronted with the question of what is a touchable area. A consistent
`principle or small set of principles greatly reduces the burden on users (Apple, 1987). Possibilities
`include: realistic button shapes, rounded rectangles, shadowed boxes, distinctive color text,
`distinctive color background, tabs on a book, or standard icons. Designers must remember that it
`can be frustrating for users who identify what they think is a selectable object, try to select it, and
`discover that it is not selectable after all. If instructions or icons are used, they should be chosen to
`be consistent with the tasks and users• expectations.
`There is no simple solution for indicating what is selectable. In some systems, all selectable
`objects appear in the same shape (Fig. 2) or color (Fig. 1 ). Once users learn this rule, all targets
`can easily be identified. Other systems place simple instructions on the screen, possibly near each
`different target. The instructions may be a simple as "Touch the desired amount .. , indicating what
`is selectable and how to select it (Fig. 3). The goal is to choose an option that works best for the
`tasks being performed, and to be consistent once the choice has been made.
`Making objects significantly different from the remainder of the screen can result in users
`ignoring the remainder of the screen. In some situations, designers may not want to make it
`obvious what is selectable: when the designer wants to force users to explore the entire screen,
`when almost everything is selectable, or when an existing image cannot be overlaid or altered. In
`these situations, the designer can either force users to select what they think is selectable, or a
`mechanism can be provided to reveal all selectable regions. A special key or a "Reveal" button may
`make all targets temporarily visible. Targets may also be shown when users try to touch a non(cid:173)
`selectable region. This method can be both frustrating and slow if there are many selectable
`regions on the screen, making a "Reveal" button preferable when possible.
`
`SCEA Ex. 1026 Page 9
`
`

`
`A new era for touchscreen applications
`
`-9-
`
`As a general guideline. if users have specific goals. making targets obvious may speed
`performance. However. if the purpose is to explore and gain general knowledge about the system.
`it may be desirable to make the targets blend in with the remainder of the screen.
`
`•Feedback
`Feedback plays an important role in every user interface. Feedback refers to indicating where
`the user is currently touching the screen. that an action could be taken. or that an action has been
`taken. Feedback provides confmnation to users that the correct actions are about to be or have
`been performed. as described in the model presented in Section 4.1.
`If users are allowed to drag their fingers before making a selection it is often advantageous to
`provide a cursor near the users• fingers showing exactly where a selection will be made if they lift
`their fingers. This is particularly important if the targets are very small. less important if targets are
`large. It is also important to indicate when a selection is possible. If the lift-off strategy (or any
`other strategy that uses the removal of the users • fingers as input) is used. then as users drag their
`fingers onto a target. users should receive feedback that a selection could be made. This could be
`visual. by inverting or flashing the target. or audible. by making a short tone. Once a selection has
`been made. feedback should indicate that an action is about to be taken. This confirmation could
`be visual or audible. Visual confirmation has the advantage that the specific action to be taken
`could be indicated (the selected target could be inverted temporarily indicating exactly what action
`will be taken). Audible feedback has the advantage that users eyes can be off the targets. however.
`audible tones are more difficult to distinguish (making them less useful for indicating the exact
`action to be taken). If audible feedback is to be used. the volume must be set carefully. If the
`system is used in a public place or in an open work area loud tones may be very annoying. if the
`tones are too soft, they may be missed. User control is a useful in this situation.
`Feedback about possible selections can often replace the cursor. In the example in Figure 4. as
`users drag their fingers across the calendar. they select a day by lifting their finger when the
`desired day is highlighted. This acts not only as a cursor. but also as an indication of possible
`selections.
`
`• Target Size
`Choosing the appropriate size for touchable regions is important when designing the user
`interface. The size of targets can depend on the physical size of the display. the number of targets
`to be presented. the relation among targets. and the type of application. The physical size of the
`display limits the size of targets for obvious reasons. In the design phase. it is often necessary to
`consider the size of the display to be used when the system is installed. The number of targets to
`be presented and the relation among these targets may influence the size of targets. If there are
`several selectable items that are closely related. and users will be selecting one or more of them.
`
`SCEA Ex. 1026 Page 10
`
`

`
`A new era for touchscreen applications
`
`-10-
`
`then they should be presented simultaneously. This allows users to view all options at one time,
`and to make an informed decision about which ones they wish to select. The type of application
`may also influence target size. If making an error may result in serious problems, targets should
`be made large enough to assure that they are accurately selected. Alternative selection strategies
`that require confirmation may also be used if errors are critical. Other possibilities include
`changing the size of the touchable regions associated with a target. If targets are not placed closely
`together, it is possible to make the touchable region larger (and possibly offset from) than the
`visible target. This allows users to make accurate selections even if they touch slightly off the
`visible target.
`Fitts' Law provides some insight when choosing target sizes. Fitts' Law relates selection times
`to the distance that must be moved and the width of the target (Fitts, 1954 ). The general form of
`the law is T =a+ b[log(2d/w)], where a and bare constants, dis the distance that must be moved,
`and w is the width of the target. This law has been demonstrated with various input devices,
`however recently a slight modification has been suggested for use with touchscreens (Sears &
`Shneiderman, 1989). This modification suggests using Fitts' Law twice, once for when users
`move their fingers to the screen, and a second time when they move their fingers on the screen. In
`general, Fitts' Law provides a method of estimating the time it will take users to select targets. It
`allows designers to predict how increasing size of the target will affect the time necessary to select
`that target It also provides insight into how the spacing between targets may relate to selection
`times (if multiple selections will be made from a single screen).
`
`• Layout of targets:
`The layout of targets refers to the location of each target on the screen and their location relative
`to each other. Targets should be placed consistently on the screen. Buttons for Quit, Back, Next,
`Cancel, etc. should be placed consistently from one screen to the next. Careful placement of
`targets on the screen can reduce or eliminate certain problems. Placing targets low on the screen
`reduces the amount of the screen obscured by users' hands when they are making selections.
`Placing targets near the edge of the screen has both advantages and disadvantages. When targets
`are near the edge of the screen, the edge acts as a barrier making it impossible for users to miss the
`target by touching too far to one side. However, placing targets near the edge of the screen may
`make selection difficult depending on how far the touchscreen is recessed into the monitor. Of
`course, the handedness of users should also be considered. Although it may be difficult, if not
`impossible, to determine the handedness of potential users when designing the interface, it may be
`possible to design the interface so neither right nor left handed users suffer.
`The spacing between targets can play an important role in the speed and accuracy of selections.
`Using basic information from Fitts' Law, we can understand the basic relationship between target
`spacing and selection rates. As targets are moved closer together, selections are faster and possibly
`
`SCEA Ex. 1026 Page 11
`
`

`
`A new era for 10uchscreen applications
`
`- 11 -
`
`higher error rates due to spacing. As targets are spaced father apart, selections will be slower, but
`error rates will decrease. The combination of spacing and target size plays an important role in
`determining both selection times and error rates. If the application requires fast, accurate selections
`with minimal attention, targets should not only be larger (as discussed earlier) but spaced farther
`apart to reduce the likelihood of incorrect selections. Target size and spacing also play an important
`role in deciding which selection strategy to use (see section on Selection Strategies).
`
`• Response time
`System response time must be quick, but not too quick. Systems must respond fast enough so
`the user knows that the computer received the input, either by immediately showing the result of
`the action or giving some feedback acknowledging the input was received. For example, when
`users select a button, the system should immediately indicate that a selection has been made, even
`if processing the input will take a few seconds. If the system is too slow indicating that a selection
`has been made, a user may attempt the selection a second time, resulting in unwanted input.
`It is possible for a system to respond too quickly. If the changes to the screen are very subtle,
`users may not notice the update if it occurs too quickly. In these situations, it is sometimes
`desirable to either provide additional feedback or slow the response slightly so users can observe
`the changes occurring.
`
`• Selection Strategies
`The following are a few of the more common selection strategies used with touchscreens.
`Each strategy is described and situations where it would be appropriate are discussed. A summary
`of when each strategy may be appropriate is provided at the end of this section.
`
`-Land-on
`When touchscreens were first introduced hardware limitations resulted in a single selection
`strategy: Iand-on. Since only the location of. the initial screen touch was available, the Iand-on
`strategy resulted in the selection of the target that was at this location (Murphy, 1987; Potter,
`Weldon & Shneiderman, 1988; Potter, Berman & Shneiderman, 1989). Using the Iand-on
`strategy, users touch the screen and the location of the touch is compared to the location of the
`targets on the screen. If the touch is on a target, that target is selected. Otherwise, users must
`remove their fingers from the screen and make another selection attempt Land-on can be used
`when targets are large enough to assure that users will not inadvertently touch an incorrect target.
`Targets approximately 2.0-2.5cm per side can be accurately selected with one attempt using the
`land-on strategy, depending on the touchscreen technology (Sears, 1990; Beringer, 1989; Hall &
`Cunningham, 1988).
`There are some applications where the targets must be selected accurately with minimal
`
`SCEA Ex. 1026 Page 12
`
`

`
`A new era for touchscreen applications
`
`-12-
`
`attention. Applications such as automobile or helicopter controls require that the selections be
`made while diverting users eyes for a minimal amount of time. When using the Iand-on strategy
`users simply look at the screen one time and touch the desired target. Since selections are made
`when users first touch the screen, additional finger movements are unimportant. The Iand-on
`strategy may be preferred for applications that require selections with