Assistive Technology Research Institute
College Misericordia - Dallas, PA 18612
 
Founded and Sponsored by the Sisters of Mercy of Dallas

 

The Elder Interface - Validation

 



The Elder Interface: Changing the Computer Experience by Accommodating the Sensory and Motor Changes of Normal Aging

Denis Anson, MS, OTR

Director of Research and Development

Assistive Technology Research Institute

Misericordia University

301 Lake St.

Dallas, PA 18636

Email: danson@misericordia.edu

Abstract

The default appearance and behaviors of the computer interface are based on the hardware limitations of the computer over a decade ago and the physical abilities of relatively young computer users.  With advancing technology, the display resolution of the typical computer has increased substantially, resulting in an increase in the visual and movement acuity required to perform tasks on the computer.  With aging, human visual and movement acuity decreases.  The default version of the interface can be difficult or impossible to use for an individual showing the changes of normal aging.

This project observed elders using computers to identify specific barriers to successful use, and then sought free and/or low-cost accommodations to the interface to make computers more usable by elders.  The resulting Elder Interface changes features of the user interface to create a more accommodating experience.  In validation, elders reported that the adapted interface was easier or about the same as the conventional interface in nearly all cases.  The single-button mouse adaptation was found to be easier to use by elder novice users, but harder to use for those with significant experience and familiarity with the two-button mouse.  None of the difficulty observed among users of the conventional interface with clicks, double-clicks, or keyboard auto-repeat were observed in those using the Elder Interface.

Elders were also observed to require substantial training in the fundamental assumptions and functions of the computer interface.  Current training does not cover the assumptions at the level needed by elders, so a different type of training may be required to fully provide the benefits of computer and internet access to elders.

Key Words:

Computer Access – Aging – Accommodation -- Interface

Introduction

Computers and internet access have been widely argued to be vital to the quality of life of elders.  Computers are touted as a means to minimize social isolation among elders who can no longer participate in community activities (Klein, 2005; Surgeon General, 1999). Telemedicine has been advocated as a means of improving health care and reducing costs for homebound elderly as well as in rural regions (Evans, Fox, Pritzl, & Halar, 1984; Evans & Jaureguy, 1982; Evans, Smith, Werkhoven, Fox, & Pritzl, 1986; Ferguson, 1998; Glueckauf et al., 1998; Nickelson, 1997; U.S. Congress, 1995).  While most of telemedicine focuses on the collection of medical information and remote monitoring of client status, telerehab seeks to deliver rehabilitation services remotely (Clark, Dawson, Scheideman-Miller, & Post, 2002).

While there are reasons to believe that computer and internet technology can be highly valuable in providing information to, and receiving health information from community dwelling elderly, in urban, suburban, and rural settings, such information provision assumes that elders are able and willing to use computer technology. However, current demographics indicate that the elderly are the least likely to be using computer technology.  In 2000, only 24% of those over age 65 had access to a computer in the home (Newburger, 2000). This is less than half the proportion of adults between 45 and 65 with computer access.  Less than 18% of those 65 and over had internet access.  By 2003, the proportion of elders with computers in the home had risen to almost 35%, but the proportion remains about half of those in the 45 to 65 group (Day, Janus, & Davis, 2003).

When asked why they did not use a computer, the economic factor (expense of the device) was seldom the reason given.  Instead, elders report that they do not have access to a computer,  find computers difficult to use, or that they are afraid they will break the machine if they do the wrong thing (Klein, 2005).  One common reason given by elders for not using computers is that they do not see any need for information from a computer.  However, after exposure to training, most elders felt that computer use is indispensable (Klein, 2005).

In examining the statistics for computer use by elders, it must be kept in mind that these are cohort statistics, not longitudinal statistics.  There is no evidence that elders become less interested in computer use over time. Rather, the population over age 65 is less likely to have used computers when working, or to have had other contact with computers than their younger cohorts. As an individual ages, the changes in physical and sensory capacity makes computer access increasingly difficult.  An individual who is using a computer may seek and find accommodations for these changes but the likelihood that an elder computer novice will successfully begin using a computer decreases with age.

Age Related Changes

Visual Changes

Normal aging produces changes in eye sight.  This is known as presbyopia. After age 45 or so, the increasing stiffness of the lens of the eye makes it difficult to focus on objects that are close, and many adults discover that they must use reading glasses or bifocals. In addition, changes to the color of the lens and to the retina result in a decreased sensitivity to light, so that greater levels of illumination are required for detailed vision. As a result, greater light levels are needed for elders to perform the same work that that a younger person can do, if, indeed, it is possible at all (Faubert, 2002; Marmor, 2006).

Physical Changes

With age, the tissues that make up the body become stiffer. (This is one of the reasons for the visual changes noted earlier.) This reduced elasticity results in restrictions in movement of the joints and muscles, and decreased flexibility in blood vessels. Because few humans are operating at maximum potential, these changes may not be noticed by an aging person for many years. In fact, a person who is functioning at a normal level into their 50s, and then begins exercising and dieting can experience significant increases in strength and mobility at a time when their maximum potential is decreasing. The maximum potential, however, does not appear to change with diet and exercise. The elder athlete simply uses more of their potential than does a younger athlete.

The articular surfaces of joints wear over time. After a long and busy life, the linings of the joints and the ligaments that hold them in alignment may begin to wear out, resulting in rough articular surfaces and poor joint alignment. This is the condition known as osteoarthritis. Osteoarthritis is not caused by disease, but simply by normal aging. However, with aging, an individual’s fingers may begin to show misalignment between the phalanges, and ulnar drift (bending toward the little-finger side of the hand) at the metacarpal-phalangeal joints. Increasing joint stiffness makes it more difficult for an elder to fully extend the arms and knees.  An elder may feel aches and pains after a busy day that would not have been felt in the past  (Bonder & Goodman, 2002; Fredericks, 1996; Norkin & Levangie, 1992).

Cognitive Changes

Along with the changes to the eyes, muscles, and joints, elders experience changes to the brain and nerves. Over time, the nerve conduction velocity decreases, so that reaction time may increase, and fine motor control and praxis decreases. In addition, the decreased quantity and sensitivity of peripheral nerve endings may reduce proprioception and kinesthesia, so that an elder may have increased difficulty identifying an object by touch, or making links between sensory modalities (Bonder & Goodman, 2002; Cotman & McGaugh, 1980; Duchek & Abreu, 1997).

The Conventional Interface

Modern mainstream computer systems universally use a “graphical user interface.” In the early days of “personal computers,” it was thought that making the computer look and behave as much like the physical world as possible would make it easier for office workers to move from typewriters and file cabinets to word processors and disk drives. While the transition, for the typical office worker, was made possible, a side effect was that many novice computer users treated word processors as typewriters and never used the advanced features that would have made their life easier.

One part of this design decision, made in the first widely available graphical computer interface (the Apple Macintosh), was to make the screen look as much like a piece of paper as possible. The original screen of the Mac was 8 1/2 inches wide, the same as a standard piece of paper, displayed black text on a white background, and had a screen resolution of 72 pixels per inch. This was important because print “point size” is measured in units of 1/72 of an inch. Thus, on the early Mac, a 12 point font would be displayed at 12 pixels high.

This equivalence of points and pixels remains an assumption of the conventional interface, as seen in the default installation settings. However, the relationship is not accurate on modern computers. The early IBM and PC Compatible computers used a monitor separate from the computer itself, and with a potentially larger screen. These Windows PCs adopted a “standard” resolution of 96 pixels to the inch rather than the 72 of the early Macintosh. This allowed for slight improvements in smoothing, and took advantage of the enhanced resolution (640x480 pixels) of the Video Graphics Adapter (VGA) monitors that came after the introduction of the Macintosh (512x342 pixels).

With advancing technology over the last decade, the size of monitors has increased, as has the resolution. From a “standard” 14 inch monitor in the early 90s, the expected monitor has expanded in size to be a 17 or 19 inch diagonal. Over half of computer users now have a screen resolution of 1024x768 or greater. A modern 15 inch laptop display may have a screen resolution of 1600x1080 or more! Yet, the display “assumptions” have not changed. The default size of fonts, icons, and other features of the interface today are the same as they were a decade ago. As a result of the increased screen resolution, a font that is displayed with the same number of pixels as on a classic monitor will now have linear dimensions that may be only half what they were when the interface was designed.

Even without the advances in computer hardware of the last decade, the Windows interface would be difficult for many elders to manage. 

While there is no universally accepted definition of “Large Print,” Industry Canada (Laughton, 2005) recommends that large print documents use 16 point typeface, upper and lower case, and san-serif fonts.  While 12 point is recognized as the “standard size for most publishing,” it is too small for many people with normal age-related changes in vision.  A 14 point font is recognized as the smallest font that can be used by many people with limited vision. However, the default size for most of the features of the Windows interface is an 8 point font. This font size, using a true point equivalent, according to Industry Canada, cannot be comfortably read by many people, even with normal vision. 

Combine the changes in technology that results in fonts being displayed below their “call size,” with the changes to vision and movement of normal aging, and the Windows interface becomes very difficult for many elders to perceive.

The Elder Interface Project

The goal of this project was to determine the features of the conventional Windows interface that pose significant barriers to elders and to identify low-cost or no-cost adaptations that would allow community-dwelling elders to use the conventional computer more readily.

Method

Design

This project was carried out in two stages.  In the development phase, community dwelling elderly were asked to perform a series of computer tasks on a number of variants of the interface.  In this phase, the actions and responses of the elders were carefully observed to identify features of the interface that presented particular problems. When a problem was identified, possible solutions were implemented and tested in the next iteration of development.

Once a stable set of adaptations was identified, the project entered the validation phase. In this phase, participants performed a set of standardized computer tasks, and were asked to rate the usability of the interface.

Participants

Participants in this study were recruited from the community and through regional senior centers.  In the majority of cases, data collection occurred at the regional senior centers using project computers, but some participants visited the Assistive Technology Research Institute testing lab.

During the development phase, 30 participants conducted 132 trials of the variations of the computer interface.  These individuals had a mean age of 69, with ages ranging from 61 to 83.  Twenty-three were female, and seven male.  Three of the participants had completed only 11 years of education, eight had education beyond high school (including one participant with a graduate degree), and the other 19 were high-school graduates.  The mean years of education for the development participants was 12.7 years.

In the validation phase of the study, an additional 77 participants were recruited to use the conventional interface, the Elder Interface, or both.  These individuals ranged in age from 61 to 90, with a mean age of 72 years.  Twenty-three of the validation participants were male, and 54 were female.  This group included eight individuals who had not completed high school, 29 who had education beyond the high school level, and 40 high school graduates.  The mean number of years of education was also 12.7 years for this group.

To participate in the study, participants had to be able to hear well enough to partake in a conversation, and have vision adequate to read 12 point printed text.  The participants could not have a history of disease or condition which affected their ability to use their hands (excluding osteoarthritis), or their vision (excluding presbyopia or other normal vision changes).

Equipment

In both the development and validation phases of the study, data were collected using computers running Windows XP, with Service Pack 2 installed.  Screen resolution varied with 1024x768 pixels on a 12 inch laptop, 1200x1024 pixels on 19 inch desktop screens, and 1600x1024 pixels on both the 15 and 17 inch laptops. Computers used had a minimum Pentium M processor operating at 1.8 Ghz, and one gigabyte of RAM.

The tasks performed by the participants required the formatting of a floppy disk (using an external USB floppy drive for the laptops), playing MP3 music files using iTunes (starting with a specific tune), viewing digital images as an electronic sideshow, entering a set of 10 checks into the Quicken 2004 money management program, selecting a scene from the DVD “The Incredibles” in PowerDVD, typing a short story in Microsoft Word, and browsing a Web site using Internet Explorer.  To control for possible transmission difficulties at remote data collection sites, the Web site was actually hosted on the test computer using the Apache web server.

Each trial session was timed using a SportLine Jumbo Timer 234 digital stopwatch.  Depending on the trial, participants used either a Microsoft USB Wheelmouse Optical or a MacAlly Mouse USB single button mouse.

The tasks to be performed were described in detailed instruction manuals that were refined during the development phase to clarify any persistent difficulties in understanding.  These instructions were given in the form of an outline, where the substeps of a task were presented at the next level of the outline.

Procedure

Prior to data collection, the researcher reviewed the informed consent document with potential participants and asked them any limitations to hearing, vision, or motor control that would exclude them from the study.  When a potential participant met inclusion guidelines, and had signed the consent form, they were asked to provide basic demographic information, including age, gender, educational level, and prior computer experience.

At the beginning of a data collection session, the participant was seated at a table with the computer before them.  They were allowed to adjust the position of the screen, keyboard, and mouse to a position where they felt comfortable.  They where shown the features of the computer they would be working on, including the floppy drive, DVD drive, and mouse (one button or two in validation sessions).  They were given the instructions: “I’d like you to use these instructions to perform a series of computer tasks.  At the highest level, you will be given a general task, such as format a floppy disk.  If you know how to do that, you just do it.  If you don’t, you move to the next level of instruction [Pointing to the instructions in the manual], and it will tell you to ‘Insert the floppy disk into the drive.’  Again, if you know how to do that, you just do it.  If not, you move to the next level of instructions.  If you get stuck, I’ll give you assistance.  You may begin.”  At this point, the timer was started, and the participant began working through the tasks.

At the end of the final step of the final task, the timer was stopped.  In the development phase, and in the many of the validation trials, the participant was given a brief break while the computer was reconfigured, and a second trial was begun.

During the development phase, many participants completed trials on multiple versions of the interface, though few tried all variants.  No participant completed more than two trials during a given session, and no session extended to more than two hours.

During the data collections sessions, the researcher carefully observed the participants to identify specific difficulties experienced in using the computer, so that alternatives could be identified.  During the development phase, these difficulties were used to create the next version of the interface.  During the validation phase, the difficulties were observed and documented.

The Elder Interface Changes

Visual Changes

As expected, most elders in the study had difficulty with the nominally 8 point font used for most features of the default interface.  A series of larger font sizes were tested in the development phase.  At sizes greater than 14 point, some participants complained that they had to scroll too much to see the information.  At 14 points, no participants complained of difficulty in reading the fonts, and while many participants had to be cued to scroll windows to find desired features, no one complained of needing to scroll too much.

While the appearance control panel allows the adjustment of many of the font sizes (and font faces) of the Windows environment, the font of the “Task Pane” of Windows XP is, by default, fixed at 8 points. 

The fourth principle of Universal design – Perceptible Information (Center for Universal Design, 1997) indicates that information should be presented in a way that makes it perceptible to the user in a wide range of conditions.  In practice, this can imply the use of adequate size, visual contrast, as well as the use of multiple sensory channels. In the Windows XP interface, task pane items are presented in a small font, but are also presented in a light to medium blue on a lighter blue background.  This color scheme provides a significant barrier for many elders and others with visual difficulties, and is not directly adjustable through the system control panel.  However, if the Windows Appearance is set to “Windows Classic” in the Display control panel, the task pane changes to black text on a white background, and follows the settings for icon labeling in the Display control panel.

One of the tasks that elders expressed interest in was web access.  While the majority of current websites present some barriers to access, we had access to a site created for a previous study that was fully compatible with W3C guidelines.  To explore the difference between conventional and elder views, a “user style-sheet” in CSS was created, which was installed in Internet Explorer 6 as part of the Elder Interface.  This change allowed websites to be displayed in 14 point, sans-serif font regardless of their default font specification.

While not strictly part of the Windows interface, the interface for Microsoft Word was also modified slightly. By default, Word opens to a window that occupies only part of the computer screen, and with a task pane on the right that occupies a portion of this window.  As a result, the document font may appear quite small.  Elders were observed working  in this reduced view, and would not attempt to enlarge the display.  To accommodate the needs of our participants while working in Word, two simple macro programs were written to be installed in the “normal.dot” file of Word.  The first of these macros (see Appendix 1), “autoexec()” forces Word to display in a full screen window, and closes the task pane.  The second macro, “autoopen(),” adjusts Word to set the zoom level to “Text width” so that the font is the largest that will allow the full width of the page to be displayed.

In spite of these changes, not all of the user interface became easier to use.  For example, the dialog box for disk formatting uses hard-coded, small font sizes that ignore the system settings.  As a result, the formatting task remained essentially the same between the two versions.  Additionally, it was discovered that many programs used in the trials have mixed ability to change font sizes.  In the Quicken 2004 program, for example, it is possible to change the font size in the check register (used to enter the checks in the study task), but the menus and surrounding cues do not change with system font changes, so they remained at 8 point (nominal) size.  In the iTunes program, used to play music files, the menu bar font ignored the system font settings, but the menus themselves did not! As a result, once a menu heading was identified in its small font, the pull-down menu was easily read.

The Windows interface does provide an alternative means of scaling the display.  The Advanced button of the Settings tab of the Display control panel allows the user to specify the “Dots Per Inch” resolution of the connected display. Using this setting, it is possible to compensate for the font size by telling the computer that the resolution of the display is even higher than it actually is.  However, in testing it was discovered that many controls of the Windows interface are specified in pixels, while the font labels may be specified in points.  As a result, the labels on some control buttons become larger than the buttons they label, and the result remains an incomplete solution.

Keyboard Changes

The changes to the keyboard behavior for the Elder Interface were relatively minor, and, for most elders, produced no discernible difference in behavior.  In the development phase of the interface, a few elders were observed holding the keys of the keyboard down when thinking, or to press keys slowly and deliberately, so that the keyboard auto-repeat produced a series of letters rather than just one.

As part of the Elder Interface adjustments, the FilterKeys option of the Accessibility Tool Panel was activated, turning off keyboard repeat.

Mouse Changes

The Elder Interface involved significant changes to the behavior of the Mouse.  During the development phase, a significant number of elders showed degenerative changes in the PIP and DIP joints of their fingers, such that their fingers tended to fall over the right mouse button rather than the left.  Due to concomitant changes in sensation, they were generally unaware of their finger position.  Additionally, the cognitive issue of right-click versus left-click (and a tendency to press the wheel of the mouse rather than the integrated buttons) posed a significant barrier to computer use.

Many elders, when attempting a “click” would scoot the mouse slightly forward as they pressed the button. The system definition of a “click” event requires that the “Mouse Down” event and the “Mouse Up” event occur within a space of four pixels. With high screen resolution, a small scoot of the mouse results in a short “drag” rather than a click.

Many elders had significant difficulty with the mouse double-click.  The default double-click requires two clicks (already difficult to produce) within 500 milliseconds, and within two pixels of each other. The slowed reaction time of elders, combined with the minute pixel size of modern displays makes double-clicking exceptionally difficult for many users.

The researcher observed that many elders had difficulty seeing the mouse cursor on the screen, and, when clicking on a control, tended to place the “center of mass” of the cursor over the control rather than the hot-spot on the tip of the arrow.  For large controls like the menus, this generally worked, but for small controls such as the close-window control, the response became erratic and frustrating.

To accommodate the needs of elders, several changes were made to the mouse and mouse behavior in the Elder Interface. The researcher replaced the standard two-button mouse with a single-button mouse (widely available for Macintosh computers, and fully compatible with PCs), and developed control strategies that do not require a right-click.  An alternative mouse-cursor set (Huge Mouse) which is much more visible, replaced the conventional mouse cursors.  Additionally, the researcher replaced the arrow-cursor with the HugeMouse cross-hairs cursor for navigation and clicking. For the Elder Interface, the “click size” and “double-click size” (the allowable distance between click events) were expanded to 32 pixels from the default 4 pixel (click) and 2 pixel (double-click) default settings. This change cannot easily be implemented directly in the Windows system, but is easily accomplished using the TweakUI utility from Microsoft. Finally, the required timing of double-clicks was extended to approximately 1500 milliseconds.  (This timing is adjustable through changes in the Windows Registry.  There is an undocumented maximum value for this setting that appears to be more than one second, but less than two seconds.  When larger values are entered, the timing does not change.)

Results

Assessment Issues

The original assessment plan was to use a series of visual-analog scales for elders to rate the level of ease of use of features of the operating system, and to do comparisons between ratings of the conventional and Elder interfaces.  This plan failed for two reasons.  First, the majority of elders in the study had difficulty with the concept of the visual analog scale.  They consistently would circle the identifiers (easy, difficult) at the ends of the scale, rather than make marks along the scale.  This forced the collapse of the visual analog scale to three ratings: easier, more difficult, and about the same.

The second difficulty with the visual analog scale was the tendency of the elders in the study to blame problems on themselves, and not on the interface.  For example, an elder may have just completed a trial during which he/she leaned far forward to get close to the computer screen and had difficulty with reading the cues on the screen, but would rate the font as “easy to read” with the verbal proviso, “if only my eyes were better.”  One participant was unable to produce successful mouse-clicks because of a tendency to “scoot” the mouse, and was only able to complete the trial when shown to hold the mouse in the air to click.  At the end of the trial, she rated the mouse “easy to use” once she had been taught the adaptive approach to clicking.

Because of these difficulties, after the first 33 participants in the validation phase, a change was made to the assessment protocol.  Behavioral observation of the participants showed marked differences in performance between the two interface versions, but participant ratings did not show a statistically significant difference because of the number of “easy” ratings in both groups.

To obtain data that matched the observed experience, participants were asked to perform a smaller range of activities on both versions of the interface, then to indicate which interface they found easier to use for the various tasks. The reduced set of tasks included the disk formatting, digital slide-show, word processing, and web access activities. These tasks were selected because they maintained the requirements for contextual menus, use of task-pane links, and keyboard use. This modified validation result will be discussed below.

Visual Issues

Participant Ratings

As expected, the visual presentation of the Elder Interface was generally preferred over the conventional interface.  Out of 80 ratings of visual differences, 56 rated the conventional interface harder to use than the Elder interface, 21 rated it as about the same, and 3 rated the conventional interface as easier to use.  These last ratings were generally in the “Formatting a floppy disk” task, where the fonts used in the tasks were identical, but the familiarity of the conventional interface as the second trial may have made the task seem easier.

Behavior Observations

During trials of the conventional interface, many participants were observed to be hunched forward over the computer to see the information on the screen.  Two participants abandoned the study during the conventional interface trials because they complained of headaches from peering at the small print. 

No participants complained of eye-strain during the Elder Interface trials.  In addition, participants were observed to sit in a more relaxed position, with less squinting at the computer screen. Regardless of the order of trials, many participants verbally commented on the difference between the interfaces.  When the Elder Interface was experienced first, they would comment, in the conventional interface trial, that they liked the other way more.  When the conventional interface was first (the minority of trials), they commented how much easier it was to see the controls in the Elder Interface.

Keyboard Issues

Participant Ratings

Of 78 ratings of keyboard ease of use, 39 found the keyboard to be about the same under the conventional interface and the Elder Interface.  In 32 cases, the conventional keyboard was rated as more difficult to use than the Elder Interface version, and in seven cases, the conventional keyboard was found to be easier to use than the Elder Interface keyboard.

Some interpretation of these results is necessary. In the Elder Interface, the keyboard behavior was changed minimally as indicated by the 50% of cases where it was rated as “about the same.”  In the seven cases where the conventional keyboard was rated as easier to use, the fact that the participant had practice with the keyboard in the Elder Interface trial was often stated as the reason.  On the other hand, in the 32 cases where the conventional keyboard was rated as more difficult to use, only three individuals were observed to have difficulty with keyboard repeats.  It seems likely that in these cases, the general difficulty of using the conventional interface (due to visual acuity and mouse action) was transferred to the keyboard.  If the individual had difficulty seeing the response to pressing a key, that might be interpreted as difficulty with the keyboard.

Behavioral Observations

For most participants, there should not have been a significant difference in the experience of the Elder Interface and the conventional interface to the keyboard.  The only change in behavior of the keyboard was turning off the keyboard repeat function.

Over the course of the trials, three participants had marked difficulty with keyboard auto-repeat producing extraneous keystrokes under the conventional interface.  No participant had difficulty with extraneous keystrokes under the Elder Interface, and no participant complained of the lack of this feature.  This included those participants who were accomplished computer users, by self report and observation.  It appears that the keyboard-repeat function, which has been standard on computers since the Apple II and TRS 80 computers of the late 1970s, is, in fact, of very little value, and should be off by default for most users.

Among those users who had never used a keyboard before (almost half of the sample), the most common error made in typing was failing to use the space bar, especially after punctuation.  In handwriting, a space is not something that you do, but is somewhere that you are. For elders who lacked keyboard experience the idea of actively inserting spaces was difficult to grasp, and would frequently be forgotten.

Among those participants who had typewriter experience, many would use the lower case “L” character for a numeral “1,” and the upper case letter “O” in place of the numeral “0.”  On early typewriters, this was, in fact, the way these numerals were produced, and this previously learned response was transferred to the computer.

Mouse Issues

Participant Ratings

Of the 80 ratings of ease of use for the mouse, 43 respondents found the mouse tasks to be harder with the conventional mouse interface than the one-button, single-click approach of the Elder Interface.  Twenty-seven respondents found the two versions to be about the same in ease of use, and 10 found the conventional mouse to be easier to use.  These 10 ratings were from individuals who had a high degree of familiarity with the two-button mouse, and were daily computer users. 

Behavioral Observations

While the visual challenges of the conventional interface were the most obvious and received the most frequent comments, the design and behavior of the computer mouse appeared to be a more significant barrier to computer access for the elders in the study.  Some of the difficulties can be mediated by system settings, but other difficulties require training in how to use the mouse.

Two primary difficulties were observed in a large portion of the participants. Many participants would hold the mouse slightly off of the table when moving it, rather than sliding it along the table (or mouse-pad) surface.  This appears to stem from the life-long conditioning against sliding dishes on the table to avoid scratching the surface.  In some cases, it was necessary to provide hand-over-hand cueing to move the mouse on the table surface.  The second and very common difficulty relates to “mouse-rowing.” The speed of cursor movement on the screen is dependent on the speed of mouse movement on the table-top, but has an acceleration curve, such that faster movement of the mouse results in greater displacement of the cursor on the screen.  But as a result, moving the mouse cursor to a desired location will often require moving the mouse to the edge of the movement space (desk-top or mouse-pad), then picking it up slightly to place it on the opposite edge of the movement surface, and moving further.  This incremental movement is known, among experienced mouse users, and rowing the cursor across the screen (by analogy with rowing a boat across a pond with a series of similar, sort movements).  While seemingly an obvious behavior, elders had substantial difficulty in grasping the concept.  They would attempt to “extend” the mouse surface by twisting and running the mouse around the periphery of the mouse pad, or would lift the mouse (often several feet into the air) and place it back in the same location. Often several minutes of training would be required to train just this one aspect of mouse movement.

In the conventional interface, participants would often ask whether they should use the right or left mouse button to perform an action.  This was especially true when using contextual menus, since they had just used the right button to call up a menu, and it seemed logical that a right click would be used to select menu items. (The instructions for the tasks consistently indicated “click the left mouse button” or “click the right mouse button” to perform a task, but this cue was inadequate for many participants.) Almost as common, among participants with beginning mouse familiarity, was the question of how to know when to use a single click or a double-click.  In the Elder Interface, neither question occurred, because this interface uses only single-clicks, and a one-button mouse.

In the conventional interface, many participants had difficulty producing successful clicks and double-clicks.  Commonly, when pressing the mouse button for a click, participants would scoot the mouse forward slightly, resulting in a “drag” rather than a “click.”  Other participants, due to hand deformities of ulnar drift, would click the right mouse button rather than the left, because their fingers naturally fell in that direction.  A few tended to click on the wheel rather than the integrated buttons, since the prominent wheel was more compelling than the smooth surface of the mouse buttons. Participants would often make several attempts before producing a successful click.  Similarly, double-clicks were very difficult for a wide range of elders under the conventional interface.  They would make multiple tries to produce clicks that were fast enough and within two pixels. 

Using the Elder Interface, no participant was observed to have difficulty producing a click, since the “scooting” behavior was accommodated by the change in drag size.  Since the Elder Interface uses a one-click approach, double-clicks were not required. (In virtually all cases, clicking once on a icon, then pressing the Enter key on the keyboard performs the same function as a double-click.) In spite of this, elders with some computer experience were observed to use double-clicks on occasion. With the modification to the double-click specification, no elder was observed to have difficulty with double-clicking. An add-on utility program (Key Mouse Genie) was used to provide right-clicks through the Control Key. Since the action did not require the use of a mouse, the participants were not told that this was a “right-click.” It was just a means of calling up contextual menus. (This can also be done through the “Menu” key of a Windows keyboard.)

A secondary benefit to the one-click interface was observed. At several points in the standard activities, participants would be asked to select an item to be acted on through another control.  For example, the digital slideshow activity instructions asked the participants to first click on a particular image to select it then activate the slideshow using the taskbar control. Under the conventional interface, a number of subjects would unintentionally double-click on the image icon, resulting in the image opening in the image viewer application. The “double-click” action seemed to be automatic when working with icons, making other activities more difficult. The single-click model of the Elder Interface did not produce these unintended activations.

Those participants who were familiar with the conventional mouse tended to rate the one-button, one-click strategy of the Elder Interface as more difficult than the conventional interface.  This is understandable, as this approach requires shifting action between the mouse and keyboard more frequently.  However, no participant found the changes in click and double-click timing and size to be difficult.

Discussion

The difficulty experienced in getting elders to rate the conventional interface as difficult to use was actually a valuable finding, in spite of the complication it created in study implementation.  Because elders will tend to blame their difficulties on themselves rather than on the computer, one cannot wait for an elder to report difficulty before intervening.  Elders are more likely to decide that they are not interested in the computer, or not smart enough to use one, especially since they are not aware of the adjustments available to them. For a computer user over the age of 60, the default configuration should be the Elder Interface.

The only feature of the Elder Interface that was rated as more difficult to use by a significant portion of the participants was the one-button mouse.  Elders who had extensive experience with a two-button, double-clicking interface rated the one-button mouse more difficult to use.  No novice computer user found the one-button mouse more difficult to use, in spite of the fact that the two-button mouse was an optical mouse, and the single-button mouse used the older, roller-ball technology.

The changes implemented in the Elder Interface are, with the exceptions of the one-button mouse and the Key Mouse Genie program, available within the operating system or at no cost from the World Wide Web (the TweakUI program needed to change mouse behavior and the HugeMouse cursor set must be downloaded, but are available at no charge).  However, the process of setting up the Elder Interface on a computer requires some computer experience, and cannot be reasonably undertaken by a computer novice.  The changes of the Elder Interface, to have a significant impact on computer use by elders, must be implemented by either a computer dealer, or, ideally, as a computer “theme” that could be selected at the time the Windows interface was first activated. 

Based on observation of the participants in this study, it is recommended that all features of the Elder Interface be implemented on any computer intended to be used by elders, with the exception of the one-button mouse.  If the computer is to be used by experienced elders, a two-button mouse is preferable.  Fortunately, with USB mice, more than one mouse can be installed and active at once, so both single- and two-button mice can be active on a computer simultaneously.

All of the features of the Elder Interface, again with the exception of  the one-button mouse, are set as “per user” settings of the interface. If a computer is to be used only by an elder, the Elder Interface can be the only configuration provided.  But if the computer is to be shared by more than one user, an individual account can be created for one or more elder users, with the features of the Elder Interface implemented.  Additional accounts for younger users or those with other specific needs will not inherit the Elder Interface settings.  Thus, sharing a computer with the Elder Interface will not negatively affect other users.

Conclusions

In the process of conducting this study, strong interest in computer technology among elders was noted in addition to,an equally strong tendency of elders to blame themselves for difficulty in using a computer. Many participants volunteered for this study, and treated it as training in how to use the computer. Some even recruited others to the “computer class.” Contrasting with this interest was the feeling among many elders that “I’m not smart enough to use the computer.” When the computer interface was hard to use, elders tended to blame themselves rather than the computer. When the screen was difficult to read, elders would describe the screen as “easy to read, if only my eyes were better.”  Since, in most cases, the physical capabilities of the elderly will not improve, this presents a substantial barrier to computer use by elders, as they are less likely to seek needed adjustments.

Over the course of this project, a number of features of the Windows interface were identified that present barriers to use by elders.  Some of these can be overcome using the accessibility features built in to Windows, but others require significant adjustments to the system.  Changes to the interface to improve access were identified.  Some of these changes helped virtually all users, while others (such as the keyboard auto-repeat behavior) were very useful to some, but unnoticed by the majority of users.

Over the course of the study it was also determined that the identified changes alone are not adequate to enable computer use by many elders.  Significant training must be provided in the way the computer works before elders will be comfortable with the interface. (Why moving the “thumb” control of a scroll-bar down moves the text of a window up, for example.)  The Windows interface commonly provides multiple ways to perform any task.  Giving novice elders more than one, consistent method of performing a task tends to confuse rather than facilitate use.

The combination of the interface changes of the Elder Interface and simple, consistent training in how to use a computer promises to substantially increase the number of elders willing and able to use computers.  The proposed benefits of telemedicine and telerehabilitation for the elderly can only be realized if the elderly are willing to use communications technologies.  The Elder Interface appears to be one important component of this willingness.

Bibliography

Bonder, B. R., & Goodman, G. (2002). Preventing occupational dysfunction secondary to aging. In C. A. Trombley & M. V. Radomski (Eds.), Occupational therapy for physical dysfunction (5th ed.). Philadelphia: Lippincott Williams & Wilkins.

Center for Universal Design. (1997, 06/25/2004). What is Universal Design- Principles of UD. Retrieved July 29, 2004, from http://www.design.ncsu.edu:8120/cud/univ_design/princ_overview.htm

Clark, P. G., Dawson, S. J., Scheideman-Miller, C., & Post, M. L. (2002). TeleRehab: Stroke teletherapy and management using two-way interactive video. Retrieved Aug. 30, 2006, from http://www.findarticles.com/p/articles/mi_qa3959/is_200206/ai_n9138658

Cotman, C. W., & McGaugh, J. L. (1980). Behavioral Neuroscience. New York: Academic Press.

Day, J. C., Janus, A., & Davis, J. (2003). Computer and Internet use in the United States: US Census Bureau.

Duchek, J. M., & Abreu, B. C. (1997). Meeting the challenge of cognitive disabilities. In C. H. Christiansen & C. Baum (Eds.), Occupational therapy: Enabling function and well-being (pp. 289-311). Thorofare, New Jersey: Slack Inc.

Evans, R. L., Fox, H. R., Pritzl, D. O., & Halar, E. M. (1984). Group treatment of physically disabled adults by telephone. Social Work in Health Care, 9(3), 77-84.

Evans, R. L., & Jaureguy, B. M. (1982). Group therapy by phone: A cognitive behavioral program for visually impaired elderly. Social Work in Health Care, 7(2), 79-90.

Evans, R. L., Smith, K. M., Werkhoven, W. S., Fox, H. R., & Pritzl, D. O. (1986). Cognitive telephone group therapy with physically disabled elderly persons. The Gerontologist, 26(1), 8-10.

Faubert, J. (2002). Visual perception and aging. Canadian Journal of Experimental Psychology, 56, 164-176.

Ferguson, T. (1998). Digital doctoring-opportunities and challenges in electronic patient-physician communication. The Journal of the American Medical Association, 280, 1361-1362.

Fredericks, C. M. (1996). Skeletal Muscle: The Somatic Effector. In C. M. Fredericks & L. K. Saladin (Eds.), Pathophysiology of the motor systems: principles and clinical presentations (pp. 59). Philadelphia: FA Davis Company.

Glueckauf, R., Whitton, J., Kain, J., Vogelgesang, S., Hudson, M., Hufford, B., et al. (1998). Home-based, videocounseling for families of rural teens with epilepsy: Program rationale and objectives. Telehealth News, 2(1), 3-5.

Klein, M. (2005). Introducing Older Persons to Benefits of Technology (Online). Oakville, ON: Sherican Institute of Technology and Advanced Research.

Laughton, M. F. (2005, 2005-09-26). Large Print. Retrieved Sept. 5, 2006, from http://www.at-links.gc.ca/guide/l3g-E.asp?Action=''&Id=l3e#top

Marmor, M. F. (2006). Normal Age-Related Vision Changes and Their Effects on Vision. Retrieved Aug. 30, 2006, from http://www.visionconnection.org/Content/ForProfessionals/PatientManagement/ManagingthePatientWithLowVision/NormalAgeRelatedVisionChangesandTheirEffectsonVision.htm?cookie%5Ftest=1

Newburger, E. C. (2000). Home Computers and Internet Use in the United States:August 2000. In U. S. Bureau (Ed.): US Census Bureau.

Nickelson, D. (1997). Telehealth Poses Opportunities and Challenges for Psychology, APA Online.

Norkin, C. C., & Levangie, P. K. (1992). Joint structure and function: a comprehensive analysis. Philadelphia: FA Davis.

Surgeon General. (1999). Depression in Older Adults. Retrieved Nov. 8, 2005, from http://www.surgeongeneral.gov/library/mentalhealth/chapter5/sec3.html#consequence

U.S. Congress, O. o. T. A. (1995). Bringing Health Care Online: The Role of Information Technologies. In O. o. T. Assessment (Ed.) (pp. 1-5): Washington, DC: U.S. Government Printing Office,.

 

Apple Computer.  Available for free download from http://www.apple.com/itunes/download/

Intuit, Inc., 2632 Marine Way, Mountain View, CA 94043. Phone: 1-800-811-8766. http://quicken.intuit.com

CyberLink USA. 46750 Fremont Blvd, Suite 200, Fremont, CA94538, U.S.A., Phone: 1 510-668-0118, http:// www.cyberlink.com

Microsoft, Corp. 1 Microsoft Way, Redmond, WA  98052-6399. http://www.microsoft.com

The Apache Software Foundation, 1901 Munsey Drive, Forest Hill, MD 21050-2747. http://www.apache.org/

Mace Group, Inc. 4601 E. Airport Drive, Ontario , CA 91761, USA, Tel: 909. 230.6888. http://www.macally.com/

VSI Systems. Phone: 310 927-4385, Web Page: http://www.vsisystems.com