|
1
|
- Jenna Melissa Shanis
- May 2, 2002
- Thesis Chair: Professor Alan Hedge PhD
- Committee Member: Professor Elizabeth Mannix
|
|
2
|
- The traditional computer input interface is inefficient for the
following reasons (Westerman, Elias, and Hedge, 2001) :
- Must switch a hand between typing and pointing activities
- Separate devices require different skill sets: typing movements for keyboards,
mostly point-and-click operations for other input devices.
- Most input devices can only be operated with one hand.
- The physical arrangement can be a risk factor in the development of
musculoskeletal disorders
- Physical area required for a keyboard and a mouse or other input device
is an inefficient use of workspace.
|
|
3
|
- Developed in the 1960s at Stanford Research Institute
- became commercially widespread in the mid-1980s when it was sold as part
of the MacIntosh™
- Advantages: can be used by either hand, causes relatively little
fatigue, is physically robust, and has convenient finger selection
buttons, and can position in different locations
- Disadvantages: associated with substantial wrist extension postures
(25-30°) and carpal tunnel pressures (up to 18.7 ± 3.8 mmHg) (Keir et
al., 1999), physical area required when using a mouse is larger than
ideal
|
|
4
|
- The touchpad is a fixed-size surface, operated by a finger, provides
positional data to the system, uses a resistive tablet to track the
point of contact between two conducting surfaces
- Touch tablet was introduced in the 1970s for drawing with a stylus
- In 1994, the Powerbook™ laptop was introduced by Apple containing the
touchpad and now has come into widespread use, integrated into laptop
computers
- Advantages: little necessity for learning, very fast and very accurate
devices with times as fast, or faster than a mouse.” (Douglas &
Mithal, 1997) In addition, a touchpad can be used with either hand and a
choice of fingers.
- Disadvantages: cannot easily perform drawing tasks. (Douglas &
Mithal, 1997), stationary nature causes wrist deviation
|
|
5
|
- enables the user to use a touchpad-like surface for keyboard input,
cursor positioning, and gesture input
- Advantages: Two-handed operation - manual advantages (increased
time-motion efficiency because there are twice as many degrees of
freedom available) and cognitive advantages (visualizing and composing
the task). Leganchuk et al (1999)
found that bimanual techniques resulted in significantly faster
performance than one-handed techniques. In addition, not limited in size
or shape and frees the constraint of touch location
|
|
6
|
- Pointing/positioning and selecting
- Dragging
- Drawing
- Domain-specific tasks: target acquisition, text selection, text editing
and entering, and continuous tracking tasks
|
|
7
|
- Musculoskeletal disorders are associated with tasks involving:
- Repetitive motions
- Sustained posture
- forceful exertions
- awkward postures
- mechanical stresses
- (Tittiranonda et al., 1999).
|
|
8
|
|
|
9
|
- Keyboard design influences hand and wrist position, typing productivity,
and preference. (Zecevic et al., 2000)
- Split angle
- Key force
- Keyboard slope
- Height of keyboard from the table and floor surfaces
|
|
10
|
- Purpose
- to explore a relatively new input technology and to determine how well
a touch surface works as a combined keyboard input, cursor positioning,
and gesture control device when compared with current input
technologies.
- The following questions were explored:
- Is keying input faster and more comfortable for a MultiTouch surface
than for a keyboard?
- Is cursor positioning faster, more comfortable, and safer for a touch
surface than for a touch pad or a mouse?
- Is text editing faster and more comfortable using MultiTouch gesture
commands than using a mouse?
|
|
11
|
- The gestures, after being learned, will save much time when compared
with using file menus and control commands to perform text editing
tasks.
- Because of the elimination of a need for force, less fatigue will be
reported by subjects using MultiTouch for data input.
- Cursor positioning will be faster and more accurate with MultiTouch than
with the mouse because it is more direct, and than the touchpad, and
because of the two finger electric field sensing technology.
|
|
12
|
- Pilot study
- 12 subjects, 6 men and 6 women
- right-handed
- Recruited from Cornell via e-mail
- 18 to 22 years old
- Tested independently in the laboratory by the same experimenter
|
|
13
|
|
|
14
|
|
|
15
|
|
|
16
|
|
|
17
|
- Data entry by copying a fixed number of numbers, 250, from a random
number table from a hard copy into the computer
- Different number groupings were used each time.
- Timed using stop-watch and total input time was recorded.
- Numberpad on two of the keyboards were compared with the keyless,
sensory MultiTouch number pad
|
|
18
|
|
|
19
|
- 3 separate, but comparable, ones were used
- approximately 1,000 words
- 10-12 grade level using a Felsch-Kincaid index
- approximately same number of each editing task
- Timed and recorded by the experimenter using a stopwatch
- Combination of Keyboard & MultiTouch Gestures were compared with
Keyboard & Mouse and Keyboard & Touchpad
|
|
20
|
- On a continuum, marked fatigue for the following:
- Right hand
- Right wrist
- Right Forearm
- Right Upper Arm
- Right Shoulder
- Neck
- Upper back
- Lower back
|
|
21
|
- Measured:
- Enjoyment and comfort of each of the input devices
- Ease of learning and performing the 2D gestures
- Frequency with which split and conventional keyboards
- Use of different types of input devices
|
|
22
|
- Completely crossed, repeated measures design
- Three conditions containing each of three tasks
- Subjects were tested in each of the three conditions in the same room
within one visit
- Condition order and tasks within each were counter-balanced and randomly
assigned to the subjects
|
|
23
|
- Subject entered the laboratory and made him/herself comfortable at
station
- Fatigue questionnaire
- MultiTouch was briefly described, demonstrated, and practiced (3 to 5
minutes long)
- Video-recording began
- Subject participated in each of three tasks (random order) with each of
the following keyboards (random order)
|
|
24
|
- Between each keyboard condition, fatigue questionnaire was completed
- Video-recording ended
- Post-trial questionnaire was given
- Subject was paid and left
|
|
25
|
- Speed (for the text editing and number input tasks)
- Number of laps completed (cursor positioning task)
- Accuracy, in percent error (cursor positioning task)
- Video recordings of the subjects’ right arm and wrist (cursor
positioning data was used)
|
|
26
|
- Peak Performance 2D system
- Collect a 10-second sample
- Digitized a random sample of 12 still images (per subject per condition)
- Points were chosen on the middle of the forearm, the wrist pivot, and on
the fourth knuckle and then connected, creating an angle
|
|
27
|
|
|
28
|
|
|
29
|
|
|
30
|
|
|
31
|
|
|
32
|
|
|
33
|
|
|
34
|
- There were no significant effects of input device on physical body
fatigue.
- There was a trend, however, that when using a touchpad, more fatigue was
reported in the right hand.
|
|
35
|
|
|
36
|
- Mean wrist extension was reduced to 32.57° compared with 34.82° when
using a mouse and 41.29° with the touchpad
- Due to differences in keyboard slope
- Feedback: visual, auditory, and haptic
|
|
37
|
- Data entry and text editing took longer with MultiTouch, but MultiTouch
was better than the touchpad for cursor positioning speed and accuracy
on a narrow track and better for speed on the wide track
- Control of accuracy necessary: wide vs. narrow
- Task Axis Crossing and registered error
|
|
38
|
- Difference in technologies: MultiTouch vs. Touchpad
- One-handed interface inferior to a two-handed - splits the compound task
into subtasks that can be performed in parallel by both hands
- MultiTouch can be used non-dominant hand or possibly bimanually
|
|
39
|
- Devices such as MultiTouch “allow an input device to express an enhanced
vocabulary of explicit actions, but the user must first learn these new
ways of touching or using the input device to fully benefit from them.”
(Hinckley & Sinclair, 1999)
- Detwiler et al. (2000) - task completion times with the MultiTouch
keyboard were significantly slower than those with the keyboard and
external mouse, but those using the MultiTouch technology after the
keyboard, fared significantly better than those exposed to the
MultiTouch technology first.
|
|
40
|
- A follow-up study was done by Hedge (2002)
- Four of the twelve subjects were used
- Completed the text passages, ending on the same one that they had
previously done with the MultiTouch keyboard in the main study
- An average time decrease of
- 38.2% (ranging from 6.7-64.5%)
- Practice may play a large role in the differences observed!
|
|
41
|
- 3D: recognition-based - such as speech, gesture, eye tracking, and
camera-based
- Problems: toolkits for application, lag
- Will have multiple input devices and will have to determine which to use
given the task or use simultaneously
- requires a deeper knowledge about how the device works
|
|
42
|
- Positive aspects:
- Sensors in MultiTouch require no mechanical intermediary, allows for a
quicker response time and direct manipulation
- Amount of feedback provided can be controlled
- Flexible form
- Unobtrusive, does not look complex and has no buttons which can easily
become damaged
- Elimination of force in order to input data
|
|
43
|
- Skill level and learning required: motor skills, memory, and perceptual
skills
- Interactive setting, whether the user is sitting, has both hands
available, and can look and focus her attention on the task at hand
|
|
44
|
|
Notes
