- Email: [email protected]
Cervical spine biomechanics and task performance during touchscreen computer operations
International Journal of Industrial Ergonomics 56 (2016) 41e50
Contents lists available at ScienceDirect
International Journal of Industrial Ergonomics journal homepage: www.elsevier.com/locate/ergon
Cervical spine biomechanics and task performance during touchscreen computer operations Boyi Hu a, b, Xiaopeng Ning c, * a
Liberty Mutual Research Institute for Safety, 71 Frankland Road, Hopkinton, MA 01748, USA Department of Environmental Health, Harvard School of Public Health, 665 Huntington Avenue, Boston, MA 02115, USA c The Ergonomics Laboratory, Department of Industrial and Management Systems Engineering, West Virginia University, P.O. Box 6070, Morgantown, WV 26506, USA b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 20 April 2016 Received in revised form 11 August 2016 Accepted 16 September 2016
The effects of different touchscreen interface designs on operators' task performance and cervical spine biomechanics were investigated in the current study. Fifteen male participants performed “Whac-aMole” type of visual target pinpointing tasks on a touchscreen monitor with different display sizes, icon sizes, icon colors and task difficulties. Participants' task performance, cervical spine biomechanics and upper extremity muscle activities were recorded and compared. Results demonstrated that an oversized desktop touchscreen monitor and small icons generated negative impacts on participants’ task performance and biomechanical measurements. Lighter icon color and more difficult task requirement generated worse task performance but had limited impact on cervical spine biomechanics. In addition, when using an oversized touchscreen monitor, the impacts of icon size and task difficulty seem to be magnified. Our results demonstrated that a more human-oriented interface design could help improve task performance and reduce neck and upper extremity injuries while operating touchscreen monitors. Relevance to industry: In this study we investigated how a number of different design factors could influence task performance as well as cervical spine biomechanics when using touchscreen monitors. Knowledge gained from the current study could help the design of future applications that involve finger touching operations on touchscreen monitors. © 2016 Elsevier B.V. All rights reserved.
Keywords: Touchscreen computer Interface design Human performance Cervical biomechanics
1. Introduction Touchscreen interfaces are becoming ubiquitous with the increasing use of touchscreen monitors and mobile devices such as smart phones and tablets. It is estimated that more than 360 million tablets will be sold worldwide by year 2016 (Young et al., 2013). Among recent laptop computer sales, the ones with touchscreen functions also accounted for a significant portion of market share (Woollacott, 2013). Previous investigations have shown that the use of computers and mobile digital devices is highly associated with the high prevalence of neck pain (Hakala et al., 2006; Berolo et al., 2011). In the general population, neck pain affects 30e50% of adults (Carroll et al., 2008) and this rate is even higher among frequent computer
* Corresponding author. E-mail addresses: [email protected] (B. Hu), [email protected]. edu (X. Ning). http://dx.doi.org/10.1016/j.ergon.2016.09.007 0169-8141/© 2016 Elsevier B.V. All rights reserved.
users (Eltayeb et al., 2009). The Bureau of Labor Statistics (BLS) reported that on average, work-related neck pain requires 11 days away from work (BLS, 2012). The cost of neck pain is also significant, one study showed treatment for neck and back problems accounted for nearly $90 billion dollars in healthcare expenditure in the United States in 2005 (Martin et al., 2008); another study estimated the direct cost related to neck pain were $185.4 million dollars in Netherlands in 1996 (Borghouts et al., 1999). Despite its high cost, the reoccurrence of neck pain is observed at 50e80% within five ^ te et al., 2008). years after its first occurrence (Co Previous studies demonstrated that the design of computer interface has a profound impact on human performance (Karwowski et al., 1994). Some of the most important design variables include: screen sizes (Jones et al., 1999), icon sizes (Huang and Lai, 2008), the color and contrast level of icons (Bzostek and Wogalter, 1999), the viewing angle and distance (Grandjean et al., 1984) and task difficulty (Orvis et al., 2008). Standards such as ISO-9241 (Ergonomics of Human-Computer Interaction) and ANSI/ HFES 100 (Human Factors Engineering of Computer Workstations)
42
B. Hu, X. Ning / International Journal of Industrial Ergonomics 56 (2016) 41e50
were developed based on the existing findings to provide guidelines for the design and development of desktop and laptop computers. However, few studies explored the effect of interface parameters on the touchscreen devices, due to its recently gained popularity. Therefore, there is a strong and urgent demand in understanding the effect of touchscreen related design features on human health and performance. In addition, previous studies have shown that desktop touchscreen usage could generate higher body discomfort (especially in the neck and shoulder region) and physical loads compared to traditional display monitors (Shin and Zhu, 2011; Kang and Shin, 2014). However, it remains unclear of how touchscreen interface parameters would influence cervical spine biomechanics. The objective of the current study was to investigate the influence of different interface designs on operators' task performance and their cervical spine biomechanics. Based on the existing literature (Jones et al., 1999; Huang and Lai, 2008; Bzostek and Wogalter, 1999; Orvis et al., 2008) changes of graphic interface parameters may alter operational performance when using nontouchscreen computers. Therefore, we suspected that changes of touchscreen interface settings would influence task performance and the cervical spine biomechanics. Specifically, we hypothesized that oversized touchscreen display, relatively smaller icon size, lower contrast level (i.e. between icons and the background display) and more difficult task will generate negative influences on users’ operational and biomechanical performance. Results of the current study may help develop future guidelines for the design of touchscreen interfaces. 2. Methods 2.1. Participants Fifteen male participants were recruited from the student population of West Virginia University and surrounding residents. Their averaged age, height and weight were 27.2 years (SD 2.6), 171.8 cm (SD 4.7) and 70.8 kg (SD 5.9) respectively. All participants were required to have at least two years of experience using touchscreen electronic devices (e.g. smartphone, tablet, etc.) and only right-handed males were recruited in order to eliminate the potential influence of sex and handedness. During the recruiting process potential participants reported their handedness, during the data collection their handedness was also verified by finishing the Edinburgh Handedness Inventory (Oldfield, 1971). Finally participants with any type of MSD that required physician visits during the past 24 months were excluded. The current research protocol was approved by the Institutional Review Board of West Virginia University. 2.2. Apparatus Bipolar surface electrodes (Bagnoli, Delsys, Boston, MA, USA) were used to collect electromyography (EMG) data from bilateral C4 paraspinal, deltoid and brachioradialis muscles with a sampling frequency of 1024 Hz. Three dimensional (3D) movement data were collected using an eight-camera optical motion sensing system (Vicon Motion System, Oxford, UK) with a sampling frequency of 100 Hz. A total of nine reflective markers were placed over the front, back and side of head (Young et al., 2012; Zhou et al., 2015), the left and right shoulders (on the most dorsal points of the clavicle bones) and the C7, T12 and S1 vertebrae (Fig. 2(a)e(c)). The Nexus software (Vicon Motion System, Oxford, UK) was used for the data collection. A custom-made computer program was built using Matlab Graphical User Interface (GUI) language (Matlab, 2011; MathWorks,
Fig. 1. An illustration of the testing program: dashed line indicates the display area for the small screen size condition while the large screen condition uses the entire display area of the screen; ‘A’ shows a large icon with darker red color, and ‘B’ shows a small icon with lighter red color. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Natick, MA, USA) to enable the testing environment. A computer workstation (Intel® Core ™ 2 Duo CPU @ 2.53 GHz, 4 GB Memory with Windows 7 installed) with a 23-inch (16:9 wide screen) touchscreen monitor was used as the testing device. 2.3. Independent variables A total of four independent variables were included in the current study, and they were: (1) screen size (SCREEN), it has two levels: 51 29 cm and 38 21 cm. These sizes were determined such that the smaller display has ~50% of the display area of the larger screen and the dimensions are both ~16:9. The smaller screen size was enabled by adjusting the display area on the same touchscreen monitor (Fig. 1). (2) icon size (ICON), it has two levels: 1.46 1.46 cm and 3.64 3.64 cm (Fig. 1). The small icon size represent roughly the size of a finger tip and the large icon size was determined through a pilot study such that the icon is easy to pinpoint without significant refined motion adjustment. (3) icon color (COLOR), it has two levels: dark red (RGB value: 140, 0, 0) and light red (RGB value: 255, 160, 160) (Fig. 1). The red color was selected based on feedbacks from a pilot study, as it tends to generate better contrast with the white background; the light red was selected so that it is significantly lighter than the dark red, yet still clear to identify from the background. (4) task difficulty (DIFFICULTY) has two levels: 1 s of target refresh rate (later referred as the “easy” condition) vs. 0.85 s of target refresh rate (later referred as the “hard” condition). The difficulty levels were determined through the same pilot study such that the high refresh rate (i.e. 0.85s) will create a clear sense of urgency and the low refresh rate (i.e. 1s) still requires participants to be fully concentrated. 2.4. Protocol Upon arrival, experimental procedures were first explained to the participants, and then a 5-min warm-up session was provided to allow participants become familiar with the tasks and computer setups. EMG electrodes were then placed to the designated locations using double-sided tapes. For the C4 paraspinal muscle, electrodes were placed bilaterally ~3 cm away from the midline of the spinal column at C4 level (Ning et al., 2015), the deltoid
B. Hu, X. Ning / International Journal of Industrial Ergonomics 56 (2016) 41e50
43
Fig. 2. (a)e(c) an illustration of locations of EMG sensors and Vicon markers; (d) a demonstration of an experimental trial.
electrodes were placed along the line from the acromion to the lateral epicondyle of the elbow, corresponding to the greatest bulge of the muscle (SENIAM, 2012); the brachioradialis electrodes were placed at the immediately distal from the cubital fossa and in line with the belly of the muscle (Beggs, 2011; Staudenmann and Taube, 2015) (Fig. 2(a)e(c)). Before the experiment, participants were required to perform two maximum voluntary contraction (MVC) trials for each muscle. EMG data recorded during MVC trials were later used to normalize EMG data collected during the task performance. When recording MVC, for C4 paraspinals, participants were secured in a custom made chair with their head flex 40 forward and performed maximum isometric head extension motion against a constant resistance (Ning et al., 2015); for deltoid, participants raised their right arm to the side, near parallel to the shoulders and performed static arm abduction against a static resistance (Dennerlein and Johnson, 2006); for brachioradialis, participants flexed elbow 90 with forearm extended to the front and upper arm by the side of the torso, held a fist under a static resistance and pushed upward. Upon finishing the MVC trials, reflective markers were placed at above described locations (Fig. 2(a)e(c)). Each participant performed a total of 32 trials (2 SCREEN 2 ICON 2 COLOR 2 DIFFICULTY 2 repetitions) in a completely randomized order. The standing distance between the participants and touchscreen was ~75% of the full arm length for each participant. They were also required to adjust the angle of the screen to the most comfortable position and this position was remained for the rest of the data collection. In each trial, participants were required to hit a total of 20 randomly appeared targets using the index finger on their dominant hand (Fig. 2(d)). Oneminute rest was provided between trials to avoid muscle fatigue.
2007); in addition, elevated neck muscle activity could result in muscle fatigue which may further lead to musculoskeletal disorders (Hu and Ning, 2015). (3) Task performance variables include the total time used to finish each trial, the number of correct clicks (hit) and the number of wrong clicks (error). 2.6. Data processing and analysis Raw EMG data were filtered with a band-pass filter (10e500 Hz) and a notch filter (60 Hz and its aliases), then fully rectified and further smoothed using a moving window filter (window width was set as 0.125 s). EMG profiles from all muscles were normalized to their corresponding maximum values recorded during MVC. Neck flexion angle was calculated as any angular deviation of the headband from the angle recorded in a forward-looking, upright standing posture, which was identified at the beginning of the data
Neck flexion angle
2.5. Dependent variables Three categories of dependent variables were included: (1) Normalized EMG (NEMG) of C4 paraspinals, deltoid and brachioradialis muscles; (2) Cervical kinematic variables include the averaged neck flexion angle (degree) in the sagittal plane, the averaged neck flexion angular velocity (degree/s) in the sagittal plane and the averaged neck angular rotation velocity (degree/s) in the transverse plane. These neck biomechanical features were selected based on previous evidence that showed that human neck is more vulnerable in forward flexed postures than in upright neutral postures (Przybyla et al., 2007), and deep and prolonged neck flexion is directly related with neck pain (Sim et al., 2006; Cagnie et al.,
Fig. 3. An illustration of the definition of the neck flexion angle.
44
B. Hu, X. Ning / International Journal of Industrial Ergonomics 56 (2016) 41e50
collection. There were 4 markers on the headband (Fig. 2(a)e(c)) and the front and back markers were used to determine the neck flexion angle (Fig. 3). Neck rotation angle was determined with the same way in the transverse plane. Angular velocities of neck flexion and rotation were also obtained. The performance data (i.e. time, hits and misses) were obtained directly from the output of the computer program. 2.7. Statistical analysis
3.2. Muscle activity Results showed using smaller icons produced significantly greater cervical extensor muscles activity (Fig. 9) and smaller right brachioradialis muscle activity (decreased from 10.6% to 8.6%) compare to using larger icons. Darker icon color resulted in higher activity on the right neck extensor muscle (increased from 9.9% to 10.3%) compare to using a lighter icon color. 3.3. Cervical kinematics
All the statistical analyses of the current study were performed using Minitab 16 (Minitab Inc., PA, USA). The ANOVA assumptions (normality of residuals, homogeneity of variance of residuals, etc.) were verified before subsequent analyses (Montgomery, 2012). Multivariate ANOVA (MANOVA) was then performed to assess the effects of main effects and their interactions on all dependent variables collectively. Significant effects were then further tested using univariate ANOVA. A criteria p-value of 0.05 was selected as the threshold for significance. 3. Results Results of the MANOVA analyses indicated significant effects of all four independent variables. Some of the interaction effects were also found to be significant. Detailed results of MANOVA and subsequent univariate ANOVA analyses are shown in Table 1. All 15 participants were right-handed and only the dominant hand was involved in the current study, therefore, NEMG results from left deltoid and left brachioradialis were not reported. 3.1. Task performance All four independent variables significantly affected participants’ operational performance. In general, larger screen size, smaller icon size, darker icon color resulted in longer total time to finish a trial, lower hits and higher errors (Figs. 4e6). The increased task difficulty by definition reduced total time (i.e. faster refresh rate will result in shorter trials) but resulted in fewer hits and more errors. Some of the interaction effects also significantly influenced performance variables. When using a larger screen, task difficulty had greater impact on the total time and the number of hits and errors (Fig. 7); similarly, icon size also had larger impact on the number of hits and errors when a larger screen was used (Fig. 8).
Results indicated that larger screen, lower task difficulty level and smaller icon size generated larger neck flexion angles (Fig. 10). Larger screen size also resulted in higher neck flexion velocity and rotation velocity (Fig. 11). In addition, smaller icon size generated higher neck flexion velocity but lower neck rotation velocity (Fig. 12). 4. Discussion The purpose of the current study was to investigate the effects of different touchscreen interface designs on the users’ cervical spine biomechanics and operational performances. Consistent with our original hypotheses, the changes of touchscreen interface settings significantly influenced most of the dependent measures. Previous studies showed that participants preferred to use larger monitors and reported better experiences (Grudin, 2001; Czerwinski et al., 2003). However, results of the current study indicated that when maintaining the same viewing distance, the 51 29 cm screen negatively affected user's performance as compare to the 38 21 cm screen. The degradation of performance might be caused by the larger viewing area and the longer distance that the arm and fingers needed to move. The current findings comply with Fitts' law, which states that the time required to rapidly switch to a specific target is a function of the ratio between the target distance and the size of target (Fitts, 1954; Soukoreff and MacKenzie, 2004). When working with a larger touchscreen monitor, participants also adopted deeper neck flexion postures (i.e. larger neck flexion angles) (Fig. 10), which have negative biomechanical influences to the cervical spine (Ming et al., 2004; Sim et al., 2006; Cagnie et al., 2007). One study suggested that more than 20 degrees of neck flexion may elevate the risk of MSDs among cervical spine (Andersen et al., 2003), in the current study the recorded neck flexion angles that were close to 20 (Fig. 10). Although the recorded neck flexion angels were relatively small, in
Table 1 The results of MANOVA and univariate ANOVA. Independent variables MANOVA ANOVA
SCREEN (S) ICON (I) COLOR (C) DIFFICULTY (D) S*I S*C S*D I*C I*D C*D S*I*C S*I*D S*C*D I*C*D S*I*C*D
p < 0.001 p < 0.001 p ¼ 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p ¼ 0.085 p < 0.001 p ¼ 0.389 p ¼ 0.144 p < 0.001 p ¼ 0.251 p ¼ 0.054 p ¼ 0.954
LC4P NEMG RC4P NEMG RD NEMG RB NEMG Hit
Error
Time
Flexion angle Flexion velocity Rotation velocity
p ¼ 0.959 p ¼ 0.002 p ¼ 0.357 p ¼ 0.074 p ¼ 0.858 p ¼ 0.588 p ¼ 0.015 N/A p ¼ 0.351 N/A N/A p < 0.001 N/A N/A N/A
p ¼ 0.005 p < 0.001 p ¼ 0.443 p < 0.001 p ¼ 0.023 p ¼ 0.530 p ¼ 0.004 N/A p ¼ 0.002 N/A N/A p ¼ 0.013 N/A N/A N/A
p < 0.001 p < 0.001 p ¼ 0.001 p < 0.001 p ¼ 0.447 p ¼ 0.068 p ¼ 0.026 N/A p < 0.001 N/A N/A p ¼ 0.339 N/A N/A N/A
p < 0.001 p < 0.001 p ¼ 0.858 p ¼ 0.014 p ¼ 0.149 p ¼ 0.639 p ¼ 0.460 N/A p ¼ 0.962 N/A N/A p ¼ 0.110 N/A N/A N/A
p ¼ 0.337 p ¼ 0.001 p ¼ 0.002 p ¼ 0.581 p ¼ 0.248 p ¼ 0.011 p ¼ 0.732 N/A p ¼ 0.706 N/A N/A p ¼ 0.361 N/A N/A N/A
p ¼ 0.001 p ¼ 0.330 p ¼ 0.761 p ¼ 0.702 p ¼ 0.018 p ¼ 0.382 p ¼ 0.045 N/A p ¼ 0.166 N/A N/A p ¼ 0.071 N/A N/A N/A
p ¼ 0.073 p < 0.001 p ¼ 0.457 p ¼ 0.962 p ¼ 0.502 p < 0.001 p ¼ 0.183 N/A p ¼ 0.153 N/A N/A p ¼ 0.467 N/A N/A N/A
p < 0.001 p < 0.001 p ¼ 0.01 p < 0.001 p < 0.001 p ¼ 0.051 p < 0.001 N/A p < 0.001 N/A N/A p ¼ 0.039 N/A N/A N/A
p < 0.001 p ¼ 0.001 p ¼ 0.448 p ¼ 0.279 p ¼ 0.496 p ¼ 0.057 p ¼ 0.527 N/A p ¼ 0.203 N/A N/A p ¼ 0.968 N/A N/A N/A
p < 0.001 p ¼ 0.011 p ¼ 0.218 p ¼ 0.927 p ¼ 0.021 p ¼ 0.860 p ¼ 0.221 N/A p ¼ 0.695 N/A N/A p ¼ 0.694 N/A N/A N/A
Note: MANOVA ¼ multivariate ANOVA; LC4P ¼ left C4 paraspinal muscle; RC4P ¼ right C4 paraspinal muscle; RD ¼ right deltoid muscle; RB ¼ right brachioradialis muscle.
B. Hu, X. Ning / International Journal of Industrial Ergonomics 56 (2016) 41e50
Light
Dark
Easy
COLOR
45
Hard DIFFICULTY
16.0
*
*
15.5
Total time (s)
15.0 14.5 ICON
16.0
*
15.5
14.0
SCREEN
*
15.0 14.5 14.0
Large
Small
Large
Small
Fig. 4. Effects of independent variables on the total time. Asterisk marks represent that the two conditions are statistically different from each other and bars indicate the corresponding standard error.
Light
Dark
Easy
COLOR
Hard DIFFICULTY
20
*
*
18
Hit (#)
16
ICON
20
*
18
14
SCREEN
*
16
14
Small
Large
Large
Small
Fig. 5. Effects of independent variables on the hit number. Asterisk marks represent that the two conditions are statistically different from each other and bars indicate the corresponding standard error.
a real occupational environment when workstation settings are not fully adjustable and/or task requirements become more difficult, much higher neck flexion angles could be observed. Partially contradictory to our hypothesis, touchscreen size did not show any significant influence on cervical extensor muscle activities. Icon size also showed significant effects on dependent variables: smaller icon caused longer time to finish a trial, less hits and more errors (Figs. 4e6). The reduced icon size may increase the difficulty of finger touching motion and therefore caused longer response time and decreased accuracy. In addition, the unique characteristics of touchscreen devices may also determined that smaller icons are less desirable; when a user operates a touchscreen monitor the floating arm may become an obstacle and block the field of view. Smaller icons are more easily covered by obstacles and more difficult to be detected by the user. Further, participants consistently maintained deeper neck flexion (Fig. 10) and higher cervical extensor muscle activity (Fig. 9) when the smaller icons were used. We
believe that participants may subconsciously adopted deeper neck flexion postures as a compensatory strategy to increase visual clarity. Such posture also resulted in higher neck extensor muscle activity to counterbalance the increased external moment. Although in the current study the sampled muscle activation levels were relatively low (often less than 15% of their maximums) and the differences were relatively small, prolonged operation of touchscreen may generate larger difference due to the development of fatigue among cervical extensor muscles. Previous literature has shown that muscle fatigue is a risk factor for MSDs (Hu and Ning, 2015); when reaching fatigue muscle often demonstrate increased activation possibly to compensate for the reduced force generating capability (Pincivero et al., 2000; Chowdhury et al., 2013; Nussbaum, 2001). Previous studies showed that during the use of nontouchscreen computers sustained low force is required among upper extremity muscles (Jorgensen et al., 1988; Zhu and Shin, 2011). Pain syndromes such as trapezius myalgia, tension neck syndrome
46
B. Hu, X. Ning / International Journal of Industrial Ergonomics 56 (2016) 41e50
Light
Dark
Easy
Hard
COLOR
DIFFICULTY
2.0
*
1.5
Error (#)
1.0 0.5 ICON
2.0
0.0
SCREEN
*
*
1.5 1.0 0.5 0.0
Large
Small
Large
Small
Fig. 6. Effects of independent variables on the error number. Asterisk marks represent that the two conditions are statistically different from each other and bars indicate the corresponding standard error.
Large Easy TotalTime
17 16
18
*
15 14 2.0 1.5
Hit
20
*
Small Easy Hard
Hard
*
*
16 14 12
Error
*
1.0 0.5 0.0 D S
Easy Hard Large
Easy Hard Small
Fig. 7. Interaction effects of DIFFICULTY and SCREEN on the total time, hit and error. “D” stands for DIFFICULTY and “S” stands for SCREEN. Asterisk marks represent that the two conditions are statistically different from each other and bars indicate the corresponding standard error.
and cervicalgia caused by low but sustained force exertions are commonly seen among prolonged PC users (Juul-Kristensen and Jensen, 2005; Blatter and Bongers, 2002; Chang et al., 2007). Finally, it remains unclear of why using larger icons generated higher neck flexion velocity, but lower neck rotation velocity. Current results showed that icon color had substantial effects on participants’ task performance. Previous studies have shown that color and contrast could significantly affect the noticeability and the time to detect an item (Laughery et al., 1993; Bzostek and Wogalter, 1999). Our results showed that lighter color significantly increased the average total time to finish a trial (Fig. 4). The task of touching on a randomly appeared icon can be divided into the icon detection phase and motion execution phase. Lighter icon color could increase the total time by expanding the response time of icon detection (Laughery et al., 1993; Bzostek and Wogalter, 1999). In contrast, the number of errors was not significantly affected by icon color (Fig. 6).
One potential explanation is that icon color and contrast mainly affects the visual input. The accuracy of touching the icon largely depends on the accuracy of human central nervous system and musculoskeletal actuator system, which may not be influenced when the position of the icon is accurately detected. As a result, icon color generated much less influence on the number of errors. Contrary to our initial hypotheses, subjects adopted larger neck flexion angles in the low difficulty condition in comparison to the high difficulty condition (Fig. 10). This finding is in agreement with one previous study (Ning et al., 2015) which also reported a less demanding task induced deeper neck flexion postures among participants when operating handheld touchscreen devices. One explanation is that when performing an easier task, participants tended to use more relaxed postures. In contrast, when performing a more demanding task, participants try to move their head back in order to keep the entire screen in their field of view. These results
B. Hu, X. Ning / International Journal of Industrial Ergonomics 56 (2016) 41e50
47
Fig. 8. Interaction effects of SCREEN and ICON on the hit and error number. “S” stands for SCREEN and “I” stands for ICON. Asterisk marks represent that the two conditions are statistically different from each other and bars indicate the corresponding standard error.
Fig. 9. Effects of ICON on the bilateral neck extensor muscles NEMG. “I” stands for ICON; LC4P stands for left C4 paraspinal muscle; RC4P stands for right C4 paraspinal muscle. Asterisk marks represent that the two conditions are statistically different from each other and bars indicate the corresponding standard error.
Easy
Hard
Neck flexion angle (°)
DIFFICULTY
Small
Large ICON
*
*
20 19 18 17 16
SCREEN
20
*
19 18 17 16
Large
Small
Fig. 10. Effects of DIFFICULTY, ICON and SCREEN on the neck flexion angle. Asterisk marks represent that the two conditions are statistically different from each other and bars indicate the corresponding standard error.
48
B. Hu, X. Ning / International Journal of Industrial Ergonomics 56 (2016) 41e50
FlexionVelocity (deg/s)
Angular velocity (deg/s)
10
10
RotationVelocity (deg/s)
* 9
9
* 8
8
7
s
7 Large
Small
Large
Small
Fig. 11. Effects of SCREEN on the neck flexion velocity and neck rotation velocity. “S” stands for SCREEN. Asterisk marks represent that the two conditions are statistically different from each other and bars indicate the corresponding standard error.
Angular velocity (deg/s)
9.0
FlexionVelocity (deg/s)
8.5
9.0
*
8.5
*
8.0
8.0
7.5
7.5
7.0
I
RotationVelocity (deg/s)
7.0 Small
Large
Small
Large
Fig. 12. Effects of ICON on the neck flexion velocity and neck rotation velocity. “I” stands for ICON. Asterisk marks represent that the two conditions are statistically different from each other and bars indicate the corresponding standard error.
suggest that task difficulty may affect the neck flexion posture in a more complicated pattern, which merits future investigation. Interesting interaction effects were found between SCREEN, ICON and DIFFICULTY, namely, when using a relatively large touchscreen monitor, the influence of ICON and DIFFICULTY on task performance is more pronounced. (Figs. 7 and 8). Human information processing theory can be used to explain this result (Lindsay and Norman, 1977; Schneider and Shiffrin, 1977). When a relative smaller display area was used, the overall demand of the task was relatively low thus more attention resources could be used to address the influence of other factors (icon size and task difficulty). On the other hand, when an over-sized screen was used, more attention is paid to monitor the increased display therefore less attention can be used to address other factors (Lutz and Huitt, 2003). These findings show that when using a relatively large touchscreen monitor, inappropriate interface designs may cause more negative influences on user performance. Several limitations of the current study need to be noted. Results
of this study demonstrated that compared with a 23 inch touchscreen, a relatively smaller display may help improve users' operational performance. The optimal size of the touchscreen monitor still merits further investigation. In addition, the task performance duration used in the current study was relatively short, future studies may explore the effects of prolonged task performance and fatigue on users’ responses. Moreover, the task performance when using non-dominant hand or both hands was not tested in the current study and warrants future investigation. 5. Conclusion In conclusion, using a relatively large touchscreen monitor with smaller size, lighter color icons and higher difficulty level may generate negative impact on users' performance and cervical kinematics. Results of the current study indicate that a more humancentered design could improve users’ neck postures and task performance when using touchscreen monitors.
B. Hu, X. Ning / International Journal of Industrial Ergonomics 56 (2016) 41e50
49
Appendix
Table 2 Detailed results for all main effects. Independent variables Levels
LC4P NEMG RC4P NEMG RD NEMG RB NEMG Hit
Error
Time
Flexion angle Flexion velocity Rotation velocity
SCREEN (S)
9.4% 2.7% 9.4% 2.8% p ¼ 0.959 9.6% 2.8% 9.2% 2.6% p ¼ 0.002 9.3% 2.6% 9.4% 2.9% p ¼ 0.357 9.3% 2.7% 9.5% 2.8% p ¼ 0.074
1.1 2.0 0.8 1.3 p ¼ 0.005 1.7 2.0 0.2 0.5 p < 0.001 1.0 1.8 0.9 1.5 p ¼ 0.443 0.7 1.2 1.2 2.0 p < 0.001
15.2 1.9 14.7 1.9 p < 0.001 15.6 1.6 14.3 2.0 p < 0.001 15.1 1.9 14.8 1.9 p ¼ 0.001 15.4 2.0 14.5 1.6 p < 0.001
18.3 7.0 17.2 7.9 p < 0.001 18.5 7.6 16.9 7.3 p < 0.001 17.7 7.1 17.7 7.9 p ¼ 0.858 17.9 7.6 17.5 7.4 p ¼ 0.014
ICON (I)
COLOR (C)
DIFFICULTY (D)
Large Mean STD Small Mean STD p-value Small Mean STD Large Mean STD p-value Light Mean STD Dark Mean STD p-value Easy Mean STD Hard Mean STD p-value
10.2% 3.3% 10.1% 3.6% p ¼ 0.337 10.3% 3.5% 9.9% 3.4% p ¼ 0.001 9.9% 3.2% 10.3% 3.7% p ¼ 0.002 10.1% 3.4% 10.2% 3.5% p ¼ 0.581
11.1% 5.9% 11.7% 5.8% p ¼ 0.001 11.4% 5.6% 11.3% 6.0% p ¼ 0.330 11.4% 5.8% 11.4% 5.8% p ¼ 0.761 11.3% 5.7% 11.4% 5.9% p ¼ 0.702
10.0% 8.0% 9.3% 7.7% p ¼ 0.073 8.6% 6.7% 10.6% 8.8% p < 0.001 9.5% 7.8% 9.8% 8.0% p ¼ 0.457 9.7% 8.1% 9.6% 7.7% p ¼ 0.962
15.6 5.4 17.0 5.0 p < 0.001 14.8 5.1 17.8 5.0 p < 0.001 16.0 5.3 16.5 5.2 p ¼ 0.01 17.4 5.0 15.2 5.2 p < 0.001
8.2 3.1 7.5 3.4 p < 0.001 7.6 3.2 8.1 3.3 p ¼ 0.001 7.8 3.2 7.9 3.3 p ¼ 0.448 7.8 3.2 7.9 3.3 p ¼ 0.279
9.2 1.8 7.1 1.2 p < 0.001 8.3 1.9 8.0 1.8 p ¼ 0.011 8.1 1.8 8.2 2.0 p ¼ 0.218 8.2 2.0 8.1 1.8 p ¼ 0.927
Note: STD¼Standard deviation; LC4P ¼ left C4 paraspinal muscle; RC4P ¼ right C4 paraspinal muscle; RD ¼ right deltoid muscle; RB ¼ right brachioradialis muscle.
References Andersen, J.H., Kaergaard, A., Mikkelsen, S., Jensen, U.F., et al., 2003. Risk factors in the onset of neck/shoulder pain in a prospective study of workers in industrial and service companies. Occup. Environ. Med. 60, 649e654. Beggs, L.A., 2011. Comparison of Muscle Activation and Kinematics during the Deadlift Using a Double-pronated and Overhand/underhand Grip. University of Kentucky, p. 87. Master’s Thesis Paper. Berolo, S., Wells, R.P., Amick III, B.C., 2011. Musculoskeletal symptoms among mobile hand-held device users and their relationship to device use: a preliminary study in a Canadian university population. Appl. Ergon. 42, 371e378. Blatter, B.M., Bongers, P.M., 2002. Duration of computer use and mouse use in relation to musculoskeletal disorders of neck or upper limb. Int. J. Ind. Ergon. 30, 295e306. BLS, 2012. Nonfatal Occupational Injuries and Illnesses Requiring Days Away from Work, 2011. Borghouts, J.A.J., Koes, B.W., Vondeling, H., Bouter, L.M., 1999. Cost-of-illness of neck pain in The Netherlands in 1996. Pain 80, 629e636. Bzostek, J.A., Wogalter, M.S., 1999. Measuring visual search time for a product warning label as a function of icon, color, column and vertical placement. In: Proceedings of the Human Factors and Ergonomics Society Annual Meeting, vol. 43, pp. 888e891. Cagnie, B., Danneels, L., Van Tiggelen, D., De Loose, V., Cambier, D., 2007. Individual and work related risk factors for neck pain among office workers: a cross sectional study. Eur. Spine J. 16, 679e686. Carroll, L.J., Hogg-Johnson, S., van der Velde, G., et al., 2008. Course and prognostic factors for neck pain in the general population: results of the bone and joint decade 2000e2010 task force on neck pain and its associated disorders. Spine 33, S75eS82. Chang, C.J., Amick Iii, B.C., Menendez, C.C., Katz, J.N., Johnson, P.W., Robertson, M., Dennerlein, J.T., 2007. Daily computer usage correlated with undergraduate students' musculoskeletal symptoms. Am. J. Ind. Med. 50, 481e488. Chowdhury, S.K., Nimbarte, A.D., Jaridi, M., Creese, R.C., 2013. Discrete wavelet transform analysis of surface electromyography for the fatigue assessment of neck and shoulder muscles. J. Electromyogr. Kinesiol. 23, 995e1003. ^ te , P., van der Velde, G., David, C.J., et al., 2008. The burden and determinants of Co neck pain in workers: results of the bone and joint decade 2000e2010 task force on neck pain and its associated disorders. Spine 33, S60eS74. Czerwinski, M., Smith, G., Regan, T., Meyers, B., et al., 2003. Toward Characterizing the Productivity Benefits of Very Large Displays. Human Computer Interaction, pp. 9e16. Dennerlein, J.T., Johnson, P.W., 2006. Different computer tasks affect the exposure of the upper extremity to biomechanical risk factors. Ergonomics 49, 45e55. Eltayeb, S., Staal, J.B., Hassan, A., De Bie, R.A., 2009. Work related risk factors for neck, shoulder and arms complaints: a cohort study among Dutch computer office workers. J. Occup. Rehabil. 19, 315e322.
Fitts, P.M., 1954. The information capacity of the human motor system in controlling the amplitude of movement. J. Exp. Psychol. 46, 381e391. Grandjean, E., Hunting, W., Nishiyama, K., 1984. Preferred VDT workstation settings, body posture and physical impairments. Appl. Ergon. 15, 99e104. Grudin, J., 2001. Partitioning digital worlds: focal and peripheral awareness in multiple monitor use. In: Proceedings of ACM CHI’92 Conference on Human Factors in Computing Systems, pp. 458e465. Hakala, P.T., Rimpela, A.H., Saarni, L.A., Salminen, J.J., 2006. Frequent computerrelated activities increase the risk of neck-shoulder and low back pain in adolescents. Eur. J. Public Health 16, 536e541. Hu, B., Ning, X., 2015. The changes of trunk motion rhythm and spinal loading during trunk flexion and extension motions caused by lumbar muscle fatigue. Ann. Biomed. Eng. 43, 2112e2119. Huang, H., Lai, H., 2008. Factors influencing the usability of icons in the LCD touchscreen. Sci. Direct 29, 339e344. Juul-Kristensen, B., Jensen, C., 2005. Self-reported workplace related ergonomic conditions as prognostic factors for musculoskeletal symptoms: the “BIT” follow up study on office workers. Occup. Environ. Med. 62, 188e194. Jones, M., Marsden, G., Mohd-Nasir, N., Boone, K., Buchanan, G., 1999. Improving web interaction on small displays. Comput. Netw. 31, 1129e1137. Jorgensen, K., Fallentin, N., Krogh-Lund, C., Jensen, B., 1988. Electromyography and fatigue during prolonged, low-level static contractions. Eur. J. Appl. Physiol. 57, 316e321. Kang, H., Shin, G., 2014. Hand usage pattern and upper body discomfort of desktop touchscreen users. Ergonomics 57. Karwowski, W., Eberts, R., Salvendy, G., Noland, S., 1994. The effects of computer interface design on human postural dynamics. Ergonomics 37, 703e724. Laughery, K.R., Young, S.L., Vaubel, K.P., Brelsfor, J.W., 1993. The noticeability of warnings on alcoholic beverage containers. J. Public Policy Mark. 12, 38e56. Lindsay, P.H., Norman, D.A., 1977. Human Information Processing: an Introduction to Psychology, second ed. Academic Press, London. Lutz, S., Huitt, W., 2003. Information Processing and Memory: Theory and Applications. In: Educational Psychology Interactive. Valdosta State University, Valdosta, GA. Retrievd [data], from. http://www.edpsycinteractive.org/papers/ infoproc.pdf. Martin, B.I., Deyo, R.A., Mirza, S.K., Turner, J.A., Comstock, B.A., Hollingworth, W., Sullivan, S.D., 2008. Expenditures and health status among adults with back and neck problems. JAMA 299, 656e664. Ming, Z., Narhi, M., Siivola, J., 2004. Neck and shoulder pain related to computer use. Pathophysiology 11, 51e56. Montgomery, D.C., 2012. Design and Analysis of Experiments, eighth ed. John Wiley and Sons, New York. Ning, X., Huang, Y., Hu, B., Nimbarte, A.D., 2015. Neck kinematics and muscle activity during mobile device operations. Int. J. Ind. Ergon. 48, 10e15. Nussbaum, M.A., 2001. Static and dynamic myoelectric measures of shoulder muscle fatigue during intermittent dynamic exertions of low to moderate
50
B. Hu, X. Ning / International Journal of Industrial Ergonomics 56 (2016) 41e50
intensity. Eur. J. Appl. Physiol. 85, 299e309. Oldfield, R.C., 1971. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 9, 97e113. Orvis, K.A., Horn, D.B., Belanich, J., 2008. The roles of task difficulty and prior videogame experience on performance and motivation in instructional videogames. Comput. Hum. Behav. 24, 2415e2433. Pincivero, D.M., Aldworth, C., Dickerson, T., Petry, C., Shultz, T., 2000. Quadriceps hamstring EMG activity during functional, closed kinetic chain exercise to fatigue. Eur. J. Appl. Physiol. 81, 504e509. Przybyla, A.S., Skrzypiec, D., Pollintine, P., Lolan, P., Adamas, M.A., 2007. Strength of the cervical spine in compression and bending. Spine 32, 1612e1620. SENIAM, 2012. Recommendations for Sensor Locations in Shoulder or Neck Muscles [online]. Available from: http://www.seniam.org. Schneider, W., Shiffrin, R.M., 1977. Controlled and automatic human information processing: I. Detection, search, and attention. Psychol. Rev. 84, 1e66. Shin, G., Zhu, X., 2011. User discomfort, work posture and muscle activity while using a touchscreen in a desktop PC setting. Ergonomics 54, 733e744. Sim, J., Lacey, R.J., Lewis, M., 2006. The impact of workplace risk factors on the occurrence of neck and upper limb pain: a general population study. BMC Public Health 6, 234.
Soukoreff, R.W., MacKenzie, I.S., 2004. Towards a standard for pointing device evaluation, perspectives on 27 years of Fitts' law research in HCI. Int. J. Hum. Comput. Stud. 61, 751e789. Staudenmann, D., Taube, W., 2015. Brachialis muscle activity can be assessed with surface electromyography. J. Electromyogr. Kinesiol. 25, 199e204. Woollacott, E., 2013. Using Windows 8 on Touchscreen Laptops Could Be a Pain in the Shoulder [press release on the internet]. Available from: http://www. theguardian.com/technology/2013/apr/09/windows-8-touchscreen-laptopspain. Young, J.G., Trudeau, M.B., Odell, D., Marinelli, K., Dennerlein, J.T., 2012. Touchscreen tablet user configurations and case-supported tilt affect head and neck flexion angles. Work 41, 81e91. Young, J.G., Trudeau, M.B., Odell, D., Marinelli, K., Dennerlein, J.T., 2013. Wrist and shoulder posture and muscle activity during touch-screen tablet use: effects of usage configuration, tablet type, and interacting hand. Work 45, 59e71. Zhou, J., Ning, X., Nimbarte, A.D., Dai, F., 2015. The assessment of material-handling strategies in dealing with sudden loading: the effect of uneven ground surface on trunk biomechanical response. Ergonomics 58, 259e267. Zhu, X., Shin, G., 2011. Shoulder and neck muscle activities during typing with articulating forearm support at different heights. Ergonomics 55, 1412e1419.
Contents lists available at ScienceDirect
International Journal of Industrial Ergonomics journal homepage: www.elsevier.com/locate/ergon
Cervical spine biomechanics and task performance during touchscreen computer operations Boyi Hu a, b, Xiaopeng Ning c, * a
Liberty Mutual Research Institute for Safety, 71 Frankland Road, Hopkinton, MA 01748, USA Department of Environmental Health, Harvard School of Public Health, 665 Huntington Avenue, Boston, MA 02115, USA c The Ergonomics Laboratory, Department of Industrial and Management Systems Engineering, West Virginia University, P.O. Box 6070, Morgantown, WV 26506, USA b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 20 April 2016 Received in revised form 11 August 2016 Accepted 16 September 2016
The effects of different touchscreen interface designs on operators' task performance and cervical spine biomechanics were investigated in the current study. Fifteen male participants performed “Whac-aMole” type of visual target pinpointing tasks on a touchscreen monitor with different display sizes, icon sizes, icon colors and task difficulties. Participants' task performance, cervical spine biomechanics and upper extremity muscle activities were recorded and compared. Results demonstrated that an oversized desktop touchscreen monitor and small icons generated negative impacts on participants’ task performance and biomechanical measurements. Lighter icon color and more difficult task requirement generated worse task performance but had limited impact on cervical spine biomechanics. In addition, when using an oversized touchscreen monitor, the impacts of icon size and task difficulty seem to be magnified. Our results demonstrated that a more human-oriented interface design could help improve task performance and reduce neck and upper extremity injuries while operating touchscreen monitors. Relevance to industry: In this study we investigated how a number of different design factors could influence task performance as well as cervical spine biomechanics when using touchscreen monitors. Knowledge gained from the current study could help the design of future applications that involve finger touching operations on touchscreen monitors. © 2016 Elsevier B.V. All rights reserved.
Keywords: Touchscreen computer Interface design Human performance Cervical biomechanics
1. Introduction Touchscreen interfaces are becoming ubiquitous with the increasing use of touchscreen monitors and mobile devices such as smart phones and tablets. It is estimated that more than 360 million tablets will be sold worldwide by year 2016 (Young et al., 2013). Among recent laptop computer sales, the ones with touchscreen functions also accounted for a significant portion of market share (Woollacott, 2013). Previous investigations have shown that the use of computers and mobile digital devices is highly associated with the high prevalence of neck pain (Hakala et al., 2006; Berolo et al., 2011). In the general population, neck pain affects 30e50% of adults (Carroll et al., 2008) and this rate is even higher among frequent computer
* Corresponding author. E-mail addresses: [email protected] (B. Hu), [email protected]. edu (X. Ning). http://dx.doi.org/10.1016/j.ergon.2016.09.007 0169-8141/© 2016 Elsevier B.V. All rights reserved.
users (Eltayeb et al., 2009). The Bureau of Labor Statistics (BLS) reported that on average, work-related neck pain requires 11 days away from work (BLS, 2012). The cost of neck pain is also significant, one study showed treatment for neck and back problems accounted for nearly $90 billion dollars in healthcare expenditure in the United States in 2005 (Martin et al., 2008); another study estimated the direct cost related to neck pain were $185.4 million dollars in Netherlands in 1996 (Borghouts et al., 1999). Despite its high cost, the reoccurrence of neck pain is observed at 50e80% within five ^ te et al., 2008). years after its first occurrence (Co Previous studies demonstrated that the design of computer interface has a profound impact on human performance (Karwowski et al., 1994). Some of the most important design variables include: screen sizes (Jones et al., 1999), icon sizes (Huang and Lai, 2008), the color and contrast level of icons (Bzostek and Wogalter, 1999), the viewing angle and distance (Grandjean et al., 1984) and task difficulty (Orvis et al., 2008). Standards such as ISO-9241 (Ergonomics of Human-Computer Interaction) and ANSI/ HFES 100 (Human Factors Engineering of Computer Workstations)
42
B. Hu, X. Ning / International Journal of Industrial Ergonomics 56 (2016) 41e50
were developed based on the existing findings to provide guidelines for the design and development of desktop and laptop computers. However, few studies explored the effect of interface parameters on the touchscreen devices, due to its recently gained popularity. Therefore, there is a strong and urgent demand in understanding the effect of touchscreen related design features on human health and performance. In addition, previous studies have shown that desktop touchscreen usage could generate higher body discomfort (especially in the neck and shoulder region) and physical loads compared to traditional display monitors (Shin and Zhu, 2011; Kang and Shin, 2014). However, it remains unclear of how touchscreen interface parameters would influence cervical spine biomechanics. The objective of the current study was to investigate the influence of different interface designs on operators' task performance and their cervical spine biomechanics. Based on the existing literature (Jones et al., 1999; Huang and Lai, 2008; Bzostek and Wogalter, 1999; Orvis et al., 2008) changes of graphic interface parameters may alter operational performance when using nontouchscreen computers. Therefore, we suspected that changes of touchscreen interface settings would influence task performance and the cervical spine biomechanics. Specifically, we hypothesized that oversized touchscreen display, relatively smaller icon size, lower contrast level (i.e. between icons and the background display) and more difficult task will generate negative influences on users’ operational and biomechanical performance. Results of the current study may help develop future guidelines for the design of touchscreen interfaces. 2. Methods 2.1. Participants Fifteen male participants were recruited from the student population of West Virginia University and surrounding residents. Their averaged age, height and weight were 27.2 years (SD 2.6), 171.8 cm (SD 4.7) and 70.8 kg (SD 5.9) respectively. All participants were required to have at least two years of experience using touchscreen electronic devices (e.g. smartphone, tablet, etc.) and only right-handed males were recruited in order to eliminate the potential influence of sex and handedness. During the recruiting process potential participants reported their handedness, during the data collection their handedness was also verified by finishing the Edinburgh Handedness Inventory (Oldfield, 1971). Finally participants with any type of MSD that required physician visits during the past 24 months were excluded. The current research protocol was approved by the Institutional Review Board of West Virginia University. 2.2. Apparatus Bipolar surface electrodes (Bagnoli, Delsys, Boston, MA, USA) were used to collect electromyography (EMG) data from bilateral C4 paraspinal, deltoid and brachioradialis muscles with a sampling frequency of 1024 Hz. Three dimensional (3D) movement data were collected using an eight-camera optical motion sensing system (Vicon Motion System, Oxford, UK) with a sampling frequency of 100 Hz. A total of nine reflective markers were placed over the front, back and side of head (Young et al., 2012; Zhou et al., 2015), the left and right shoulders (on the most dorsal points of the clavicle bones) and the C7, T12 and S1 vertebrae (Fig. 2(a)e(c)). The Nexus software (Vicon Motion System, Oxford, UK) was used for the data collection. A custom-made computer program was built using Matlab Graphical User Interface (GUI) language (Matlab, 2011; MathWorks,
Fig. 1. An illustration of the testing program: dashed line indicates the display area for the small screen size condition while the large screen condition uses the entire display area of the screen; ‘A’ shows a large icon with darker red color, and ‘B’ shows a small icon with lighter red color. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Natick, MA, USA) to enable the testing environment. A computer workstation (Intel® Core ™ 2 Duo CPU @ 2.53 GHz, 4 GB Memory with Windows 7 installed) with a 23-inch (16:9 wide screen) touchscreen monitor was used as the testing device. 2.3. Independent variables A total of four independent variables were included in the current study, and they were: (1) screen size (SCREEN), it has two levels: 51 29 cm and 38 21 cm. These sizes were determined such that the smaller display has ~50% of the display area of the larger screen and the dimensions are both ~16:9. The smaller screen size was enabled by adjusting the display area on the same touchscreen monitor (Fig. 1). (2) icon size (ICON), it has two levels: 1.46 1.46 cm and 3.64 3.64 cm (Fig. 1). The small icon size represent roughly the size of a finger tip and the large icon size was determined through a pilot study such that the icon is easy to pinpoint without significant refined motion adjustment. (3) icon color (COLOR), it has two levels: dark red (RGB value: 140, 0, 0) and light red (RGB value: 255, 160, 160) (Fig. 1). The red color was selected based on feedbacks from a pilot study, as it tends to generate better contrast with the white background; the light red was selected so that it is significantly lighter than the dark red, yet still clear to identify from the background. (4) task difficulty (DIFFICULTY) has two levels: 1 s of target refresh rate (later referred as the “easy” condition) vs. 0.85 s of target refresh rate (later referred as the “hard” condition). The difficulty levels were determined through the same pilot study such that the high refresh rate (i.e. 0.85s) will create a clear sense of urgency and the low refresh rate (i.e. 1s) still requires participants to be fully concentrated. 2.4. Protocol Upon arrival, experimental procedures were first explained to the participants, and then a 5-min warm-up session was provided to allow participants become familiar with the tasks and computer setups. EMG electrodes were then placed to the designated locations using double-sided tapes. For the C4 paraspinal muscle, electrodes were placed bilaterally ~3 cm away from the midline of the spinal column at C4 level (Ning et al., 2015), the deltoid
B. Hu, X. Ning / International Journal of Industrial Ergonomics 56 (2016) 41e50
43
Fig. 2. (a)e(c) an illustration of locations of EMG sensors and Vicon markers; (d) a demonstration of an experimental trial.
electrodes were placed along the line from the acromion to the lateral epicondyle of the elbow, corresponding to the greatest bulge of the muscle (SENIAM, 2012); the brachioradialis electrodes were placed at the immediately distal from the cubital fossa and in line with the belly of the muscle (Beggs, 2011; Staudenmann and Taube, 2015) (Fig. 2(a)e(c)). Before the experiment, participants were required to perform two maximum voluntary contraction (MVC) trials for each muscle. EMG data recorded during MVC trials were later used to normalize EMG data collected during the task performance. When recording MVC, for C4 paraspinals, participants were secured in a custom made chair with their head flex 40 forward and performed maximum isometric head extension motion against a constant resistance (Ning et al., 2015); for deltoid, participants raised their right arm to the side, near parallel to the shoulders and performed static arm abduction against a static resistance (Dennerlein and Johnson, 2006); for brachioradialis, participants flexed elbow 90 with forearm extended to the front and upper arm by the side of the torso, held a fist under a static resistance and pushed upward. Upon finishing the MVC trials, reflective markers were placed at above described locations (Fig. 2(a)e(c)). Each participant performed a total of 32 trials (2 SCREEN 2 ICON 2 COLOR 2 DIFFICULTY 2 repetitions) in a completely randomized order. The standing distance between the participants and touchscreen was ~75% of the full arm length for each participant. They were also required to adjust the angle of the screen to the most comfortable position and this position was remained for the rest of the data collection. In each trial, participants were required to hit a total of 20 randomly appeared targets using the index finger on their dominant hand (Fig. 2(d)). Oneminute rest was provided between trials to avoid muscle fatigue.
2007); in addition, elevated neck muscle activity could result in muscle fatigue which may further lead to musculoskeletal disorders (Hu and Ning, 2015). (3) Task performance variables include the total time used to finish each trial, the number of correct clicks (hit) and the number of wrong clicks (error). 2.6. Data processing and analysis Raw EMG data were filtered with a band-pass filter (10e500 Hz) and a notch filter (60 Hz and its aliases), then fully rectified and further smoothed using a moving window filter (window width was set as 0.125 s). EMG profiles from all muscles were normalized to their corresponding maximum values recorded during MVC. Neck flexion angle was calculated as any angular deviation of the headband from the angle recorded in a forward-looking, upright standing posture, which was identified at the beginning of the data
Neck flexion angle
2.5. Dependent variables Three categories of dependent variables were included: (1) Normalized EMG (NEMG) of C4 paraspinals, deltoid and brachioradialis muscles; (2) Cervical kinematic variables include the averaged neck flexion angle (degree) in the sagittal plane, the averaged neck flexion angular velocity (degree/s) in the sagittal plane and the averaged neck angular rotation velocity (degree/s) in the transverse plane. These neck biomechanical features were selected based on previous evidence that showed that human neck is more vulnerable in forward flexed postures than in upright neutral postures (Przybyla et al., 2007), and deep and prolonged neck flexion is directly related with neck pain (Sim et al., 2006; Cagnie et al.,
Fig. 3. An illustration of the definition of the neck flexion angle.
44
B. Hu, X. Ning / International Journal of Industrial Ergonomics 56 (2016) 41e50
collection. There were 4 markers on the headband (Fig. 2(a)e(c)) and the front and back markers were used to determine the neck flexion angle (Fig. 3). Neck rotation angle was determined with the same way in the transverse plane. Angular velocities of neck flexion and rotation were also obtained. The performance data (i.e. time, hits and misses) were obtained directly from the output of the computer program. 2.7. Statistical analysis
3.2. Muscle activity Results showed using smaller icons produced significantly greater cervical extensor muscles activity (Fig. 9) and smaller right brachioradialis muscle activity (decreased from 10.6% to 8.6%) compare to using larger icons. Darker icon color resulted in higher activity on the right neck extensor muscle (increased from 9.9% to 10.3%) compare to using a lighter icon color. 3.3. Cervical kinematics
All the statistical analyses of the current study were performed using Minitab 16 (Minitab Inc., PA, USA). The ANOVA assumptions (normality of residuals, homogeneity of variance of residuals, etc.) were verified before subsequent analyses (Montgomery, 2012). Multivariate ANOVA (MANOVA) was then performed to assess the effects of main effects and their interactions on all dependent variables collectively. Significant effects were then further tested using univariate ANOVA. A criteria p-value of 0.05 was selected as the threshold for significance. 3. Results Results of the MANOVA analyses indicated significant effects of all four independent variables. Some of the interaction effects were also found to be significant. Detailed results of MANOVA and subsequent univariate ANOVA analyses are shown in Table 1. All 15 participants were right-handed and only the dominant hand was involved in the current study, therefore, NEMG results from left deltoid and left brachioradialis were not reported. 3.1. Task performance All four independent variables significantly affected participants’ operational performance. In general, larger screen size, smaller icon size, darker icon color resulted in longer total time to finish a trial, lower hits and higher errors (Figs. 4e6). The increased task difficulty by definition reduced total time (i.e. faster refresh rate will result in shorter trials) but resulted in fewer hits and more errors. Some of the interaction effects also significantly influenced performance variables. When using a larger screen, task difficulty had greater impact on the total time and the number of hits and errors (Fig. 7); similarly, icon size also had larger impact on the number of hits and errors when a larger screen was used (Fig. 8).
Results indicated that larger screen, lower task difficulty level and smaller icon size generated larger neck flexion angles (Fig. 10). Larger screen size also resulted in higher neck flexion velocity and rotation velocity (Fig. 11). In addition, smaller icon size generated higher neck flexion velocity but lower neck rotation velocity (Fig. 12). 4. Discussion The purpose of the current study was to investigate the effects of different touchscreen interface designs on the users’ cervical spine biomechanics and operational performances. Consistent with our original hypotheses, the changes of touchscreen interface settings significantly influenced most of the dependent measures. Previous studies showed that participants preferred to use larger monitors and reported better experiences (Grudin, 2001; Czerwinski et al., 2003). However, results of the current study indicated that when maintaining the same viewing distance, the 51 29 cm screen negatively affected user's performance as compare to the 38 21 cm screen. The degradation of performance might be caused by the larger viewing area and the longer distance that the arm and fingers needed to move. The current findings comply with Fitts' law, which states that the time required to rapidly switch to a specific target is a function of the ratio between the target distance and the size of target (Fitts, 1954; Soukoreff and MacKenzie, 2004). When working with a larger touchscreen monitor, participants also adopted deeper neck flexion postures (i.e. larger neck flexion angles) (Fig. 10), which have negative biomechanical influences to the cervical spine (Ming et al., 2004; Sim et al., 2006; Cagnie et al., 2007). One study suggested that more than 20 degrees of neck flexion may elevate the risk of MSDs among cervical spine (Andersen et al., 2003), in the current study the recorded neck flexion angles that were close to 20 (Fig. 10). Although the recorded neck flexion angels were relatively small, in
Table 1 The results of MANOVA and univariate ANOVA. Independent variables MANOVA ANOVA
SCREEN (S) ICON (I) COLOR (C) DIFFICULTY (D) S*I S*C S*D I*C I*D C*D S*I*C S*I*D S*C*D I*C*D S*I*C*D
p < 0.001 p < 0.001 p ¼ 0.001 p < 0.001 p < 0.001 p < 0.001 p < 0.001 p ¼ 0.085 p < 0.001 p ¼ 0.389 p ¼ 0.144 p < 0.001 p ¼ 0.251 p ¼ 0.054 p ¼ 0.954
LC4P NEMG RC4P NEMG RD NEMG RB NEMG Hit
Error
Time
Flexion angle Flexion velocity Rotation velocity
p ¼ 0.959 p ¼ 0.002 p ¼ 0.357 p ¼ 0.074 p ¼ 0.858 p ¼ 0.588 p ¼ 0.015 N/A p ¼ 0.351 N/A N/A p < 0.001 N/A N/A N/A
p ¼ 0.005 p < 0.001 p ¼ 0.443 p < 0.001 p ¼ 0.023 p ¼ 0.530 p ¼ 0.004 N/A p ¼ 0.002 N/A N/A p ¼ 0.013 N/A N/A N/A
p < 0.001 p < 0.001 p ¼ 0.001 p < 0.001 p ¼ 0.447 p ¼ 0.068 p ¼ 0.026 N/A p < 0.001 N/A N/A p ¼ 0.339 N/A N/A N/A
p < 0.001 p < 0.001 p ¼ 0.858 p ¼ 0.014 p ¼ 0.149 p ¼ 0.639 p ¼ 0.460 N/A p ¼ 0.962 N/A N/A p ¼ 0.110 N/A N/A N/A
p ¼ 0.337 p ¼ 0.001 p ¼ 0.002 p ¼ 0.581 p ¼ 0.248 p ¼ 0.011 p ¼ 0.732 N/A p ¼ 0.706 N/A N/A p ¼ 0.361 N/A N/A N/A
p ¼ 0.001 p ¼ 0.330 p ¼ 0.761 p ¼ 0.702 p ¼ 0.018 p ¼ 0.382 p ¼ 0.045 N/A p ¼ 0.166 N/A N/A p ¼ 0.071 N/A N/A N/A
p ¼ 0.073 p < 0.001 p ¼ 0.457 p ¼ 0.962 p ¼ 0.502 p < 0.001 p ¼ 0.183 N/A p ¼ 0.153 N/A N/A p ¼ 0.467 N/A N/A N/A
p < 0.001 p < 0.001 p ¼ 0.01 p < 0.001 p < 0.001 p ¼ 0.051 p < 0.001 N/A p < 0.001 N/A N/A p ¼ 0.039 N/A N/A N/A
p < 0.001 p ¼ 0.001 p ¼ 0.448 p ¼ 0.279 p ¼ 0.496 p ¼ 0.057 p ¼ 0.527 N/A p ¼ 0.203 N/A N/A p ¼ 0.968 N/A N/A N/A
p < 0.001 p ¼ 0.011 p ¼ 0.218 p ¼ 0.927 p ¼ 0.021 p ¼ 0.860 p ¼ 0.221 N/A p ¼ 0.695 N/A N/A p ¼ 0.694 N/A N/A N/A
Note: MANOVA ¼ multivariate ANOVA; LC4P ¼ left C4 paraspinal muscle; RC4P ¼ right C4 paraspinal muscle; RD ¼ right deltoid muscle; RB ¼ right brachioradialis muscle.
B. Hu, X. Ning / International Journal of Industrial Ergonomics 56 (2016) 41e50
Light
Dark
Easy
COLOR
45
Hard DIFFICULTY
16.0
*
*
15.5
Total time (s)
15.0 14.5 ICON
16.0
*
15.5
14.0
SCREEN
*
15.0 14.5 14.0
Large
Small
Large
Small
Fig. 4. Effects of independent variables on the total time. Asterisk marks represent that the two conditions are statistically different from each other and bars indicate the corresponding standard error.
Light
Dark
Easy
COLOR
Hard DIFFICULTY
20
*
*
18
Hit (#)
16
ICON
20
*
18
14
SCREEN
*
16
14
Small
Large
Large
Small
Fig. 5. Effects of independent variables on the hit number. Asterisk marks represent that the two conditions are statistically different from each other and bars indicate the corresponding standard error.
a real occupational environment when workstation settings are not fully adjustable and/or task requirements become more difficult, much higher neck flexion angles could be observed. Partially contradictory to our hypothesis, touchscreen size did not show any significant influence on cervical extensor muscle activities. Icon size also showed significant effects on dependent variables: smaller icon caused longer time to finish a trial, less hits and more errors (Figs. 4e6). The reduced icon size may increase the difficulty of finger touching motion and therefore caused longer response time and decreased accuracy. In addition, the unique characteristics of touchscreen devices may also determined that smaller icons are less desirable; when a user operates a touchscreen monitor the floating arm may become an obstacle and block the field of view. Smaller icons are more easily covered by obstacles and more difficult to be detected by the user. Further, participants consistently maintained deeper neck flexion (Fig. 10) and higher cervical extensor muscle activity (Fig. 9) when the smaller icons were used. We
believe that participants may subconsciously adopted deeper neck flexion postures as a compensatory strategy to increase visual clarity. Such posture also resulted in higher neck extensor muscle activity to counterbalance the increased external moment. Although in the current study the sampled muscle activation levels were relatively low (often less than 15% of their maximums) and the differences were relatively small, prolonged operation of touchscreen may generate larger difference due to the development of fatigue among cervical extensor muscles. Previous literature has shown that muscle fatigue is a risk factor for MSDs (Hu and Ning, 2015); when reaching fatigue muscle often demonstrate increased activation possibly to compensate for the reduced force generating capability (Pincivero et al., 2000; Chowdhury et al., 2013; Nussbaum, 2001). Previous studies showed that during the use of nontouchscreen computers sustained low force is required among upper extremity muscles (Jorgensen et al., 1988; Zhu and Shin, 2011). Pain syndromes such as trapezius myalgia, tension neck syndrome
46
B. Hu, X. Ning / International Journal of Industrial Ergonomics 56 (2016) 41e50
Light
Dark
Easy
Hard
COLOR
DIFFICULTY
2.0
*
1.5
Error (#)
1.0 0.5 ICON
2.0
0.0
SCREEN
*
*
1.5 1.0 0.5 0.0
Large
Small
Large
Small
Fig. 6. Effects of independent variables on the error number. Asterisk marks represent that the two conditions are statistically different from each other and bars indicate the corresponding standard error.
Large Easy TotalTime
17 16
18
*
15 14 2.0 1.5
Hit
20
*
Small Easy Hard
Hard
*
*
16 14 12
Error
*
1.0 0.5 0.0 D S
Easy Hard Large
Easy Hard Small
Fig. 7. Interaction effects of DIFFICULTY and SCREEN on the total time, hit and error. “D” stands for DIFFICULTY and “S” stands for SCREEN. Asterisk marks represent that the two conditions are statistically different from each other and bars indicate the corresponding standard error.
and cervicalgia caused by low but sustained force exertions are commonly seen among prolonged PC users (Juul-Kristensen and Jensen, 2005; Blatter and Bongers, 2002; Chang et al., 2007). Finally, it remains unclear of why using larger icons generated higher neck flexion velocity, but lower neck rotation velocity. Current results showed that icon color had substantial effects on participants’ task performance. Previous studies have shown that color and contrast could significantly affect the noticeability and the time to detect an item (Laughery et al., 1993; Bzostek and Wogalter, 1999). Our results showed that lighter color significantly increased the average total time to finish a trial (Fig. 4). The task of touching on a randomly appeared icon can be divided into the icon detection phase and motion execution phase. Lighter icon color could increase the total time by expanding the response time of icon detection (Laughery et al., 1993; Bzostek and Wogalter, 1999). In contrast, the number of errors was not significantly affected by icon color (Fig. 6).
One potential explanation is that icon color and contrast mainly affects the visual input. The accuracy of touching the icon largely depends on the accuracy of human central nervous system and musculoskeletal actuator system, which may not be influenced when the position of the icon is accurately detected. As a result, icon color generated much less influence on the number of errors. Contrary to our initial hypotheses, subjects adopted larger neck flexion angles in the low difficulty condition in comparison to the high difficulty condition (Fig. 10). This finding is in agreement with one previous study (Ning et al., 2015) which also reported a less demanding task induced deeper neck flexion postures among participants when operating handheld touchscreen devices. One explanation is that when performing an easier task, participants tended to use more relaxed postures. In contrast, when performing a more demanding task, participants try to move their head back in order to keep the entire screen in their field of view. These results
B. Hu, X. Ning / International Journal of Industrial Ergonomics 56 (2016) 41e50
47
Fig. 8. Interaction effects of SCREEN and ICON on the hit and error number. “S” stands for SCREEN and “I” stands for ICON. Asterisk marks represent that the two conditions are statistically different from each other and bars indicate the corresponding standard error.
Fig. 9. Effects of ICON on the bilateral neck extensor muscles NEMG. “I” stands for ICON; LC4P stands for left C4 paraspinal muscle; RC4P stands for right C4 paraspinal muscle. Asterisk marks represent that the two conditions are statistically different from each other and bars indicate the corresponding standard error.
Easy
Hard
Neck flexion angle (°)
DIFFICULTY
Small
Large ICON
*
*
20 19 18 17 16
SCREEN
20
*
19 18 17 16
Large
Small
Fig. 10. Effects of DIFFICULTY, ICON and SCREEN on the neck flexion angle. Asterisk marks represent that the two conditions are statistically different from each other and bars indicate the corresponding standard error.
48
B. Hu, X. Ning / International Journal of Industrial Ergonomics 56 (2016) 41e50
FlexionVelocity (deg/s)
Angular velocity (deg/s)
10
10
RotationVelocity (deg/s)
* 9
9
* 8
8
7
s
7 Large
Small
Large
Small
Fig. 11. Effects of SCREEN on the neck flexion velocity and neck rotation velocity. “S” stands for SCREEN. Asterisk marks represent that the two conditions are statistically different from each other and bars indicate the corresponding standard error.
Angular velocity (deg/s)
9.0
FlexionVelocity (deg/s)
8.5
9.0
*
8.5
*
8.0
8.0
7.5
7.5
7.0
I
RotationVelocity (deg/s)
7.0 Small
Large
Small
Large
Fig. 12. Effects of ICON on the neck flexion velocity and neck rotation velocity. “I” stands for ICON. Asterisk marks represent that the two conditions are statistically different from each other and bars indicate the corresponding standard error.
suggest that task difficulty may affect the neck flexion posture in a more complicated pattern, which merits future investigation. Interesting interaction effects were found between SCREEN, ICON and DIFFICULTY, namely, when using a relatively large touchscreen monitor, the influence of ICON and DIFFICULTY on task performance is more pronounced. (Figs. 7 and 8). Human information processing theory can be used to explain this result (Lindsay and Norman, 1977; Schneider and Shiffrin, 1977). When a relative smaller display area was used, the overall demand of the task was relatively low thus more attention resources could be used to address the influence of other factors (icon size and task difficulty). On the other hand, when an over-sized screen was used, more attention is paid to monitor the increased display therefore less attention can be used to address other factors (Lutz and Huitt, 2003). These findings show that when using a relatively large touchscreen monitor, inappropriate interface designs may cause more negative influences on user performance. Several limitations of the current study need to be noted. Results
of this study demonstrated that compared with a 23 inch touchscreen, a relatively smaller display may help improve users' operational performance. The optimal size of the touchscreen monitor still merits further investigation. In addition, the task performance duration used in the current study was relatively short, future studies may explore the effects of prolonged task performance and fatigue on users’ responses. Moreover, the task performance when using non-dominant hand or both hands was not tested in the current study and warrants future investigation. 5. Conclusion In conclusion, using a relatively large touchscreen monitor with smaller size, lighter color icons and higher difficulty level may generate negative impact on users' performance and cervical kinematics. Results of the current study indicate that a more humancentered design could improve users’ neck postures and task performance when using touchscreen monitors.
B. Hu, X. Ning / International Journal of Industrial Ergonomics 56 (2016) 41e50
49
Appendix
Table 2 Detailed results for all main effects. Independent variables Levels
LC4P NEMG RC4P NEMG RD NEMG RB NEMG Hit
Error
Time
Flexion angle Flexion velocity Rotation velocity
SCREEN (S)
9.4% 2.7% 9.4% 2.8% p ¼ 0.959 9.6% 2.8% 9.2% 2.6% p ¼ 0.002 9.3% 2.6% 9.4% 2.9% p ¼ 0.357 9.3% 2.7% 9.5% 2.8% p ¼ 0.074
1.1 2.0 0.8 1.3 p ¼ 0.005 1.7 2.0 0.2 0.5 p < 0.001 1.0 1.8 0.9 1.5 p ¼ 0.443 0.7 1.2 1.2 2.0 p < 0.001
15.2 1.9 14.7 1.9 p < 0.001 15.6 1.6 14.3 2.0 p < 0.001 15.1 1.9 14.8 1.9 p ¼ 0.001 15.4 2.0 14.5 1.6 p < 0.001
18.3 7.0 17.2 7.9 p < 0.001 18.5 7.6 16.9 7.3 p < 0.001 17.7 7.1 17.7 7.9 p ¼ 0.858 17.9 7.6 17.5 7.4 p ¼ 0.014
ICON (I)
COLOR (C)
DIFFICULTY (D)
Large Mean STD Small Mean STD p-value Small Mean STD Large Mean STD p-value Light Mean STD Dark Mean STD p-value Easy Mean STD Hard Mean STD p-value
10.2% 3.3% 10.1% 3.6% p ¼ 0.337 10.3% 3.5% 9.9% 3.4% p ¼ 0.001 9.9% 3.2% 10.3% 3.7% p ¼ 0.002 10.1% 3.4% 10.2% 3.5% p ¼ 0.581
11.1% 5.9% 11.7% 5.8% p ¼ 0.001 11.4% 5.6% 11.3% 6.0% p ¼ 0.330 11.4% 5.8% 11.4% 5.8% p ¼ 0.761 11.3% 5.7% 11.4% 5.9% p ¼ 0.702
10.0% 8.0% 9.3% 7.7% p ¼ 0.073 8.6% 6.7% 10.6% 8.8% p < 0.001 9.5% 7.8% 9.8% 8.0% p ¼ 0.457 9.7% 8.1% 9.6% 7.7% p ¼ 0.962
15.6 5.4 17.0 5.0 p < 0.001 14.8 5.1 17.8 5.0 p < 0.001 16.0 5.3 16.5 5.2 p ¼ 0.01 17.4 5.0 15.2 5.2 p < 0.001
8.2 3.1 7.5 3.4 p < 0.001 7.6 3.2 8.1 3.3 p ¼ 0.001 7.8 3.2 7.9 3.3 p ¼ 0.448 7.8 3.2 7.9 3.3 p ¼ 0.279
9.2 1.8 7.1 1.2 p < 0.001 8.3 1.9 8.0 1.8 p ¼ 0.011 8.1 1.8 8.2 2.0 p ¼ 0.218 8.2 2.0 8.1 1.8 p ¼ 0.927
Note: STD¼Standard deviation; LC4P ¼ left C4 paraspinal muscle; RC4P ¼ right C4 paraspinal muscle; RD ¼ right deltoid muscle; RB ¼ right brachioradialis muscle.
References Andersen, J.H., Kaergaard, A., Mikkelsen, S., Jensen, U.F., et al., 2003. Risk factors in the onset of neck/shoulder pain in a prospective study of workers in industrial and service companies. Occup. Environ. Med. 60, 649e654. Beggs, L.A., 2011. Comparison of Muscle Activation and Kinematics during the Deadlift Using a Double-pronated and Overhand/underhand Grip. University of Kentucky, p. 87. Master’s Thesis Paper. Berolo, S., Wells, R.P., Amick III, B.C., 2011. Musculoskeletal symptoms among mobile hand-held device users and their relationship to device use: a preliminary study in a Canadian university population. Appl. Ergon. 42, 371e378. Blatter, B.M., Bongers, P.M., 2002. Duration of computer use and mouse use in relation to musculoskeletal disorders of neck or upper limb. Int. J. Ind. Ergon. 30, 295e306. BLS, 2012. Nonfatal Occupational Injuries and Illnesses Requiring Days Away from Work, 2011. Borghouts, J.A.J., Koes, B.W., Vondeling, H., Bouter, L.M., 1999. Cost-of-illness of neck pain in The Netherlands in 1996. Pain 80, 629e636. Bzostek, J.A., Wogalter, M.S., 1999. Measuring visual search time for a product warning label as a function of icon, color, column and vertical placement. In: Proceedings of the Human Factors and Ergonomics Society Annual Meeting, vol. 43, pp. 888e891. Cagnie, B., Danneels, L., Van Tiggelen, D., De Loose, V., Cambier, D., 2007. Individual and work related risk factors for neck pain among office workers: a cross sectional study. Eur. Spine J. 16, 679e686. Carroll, L.J., Hogg-Johnson, S., van der Velde, G., et al., 2008. Course and prognostic factors for neck pain in the general population: results of the bone and joint decade 2000e2010 task force on neck pain and its associated disorders. Spine 33, S75eS82. Chang, C.J., Amick Iii, B.C., Menendez, C.C., Katz, J.N., Johnson, P.W., Robertson, M., Dennerlein, J.T., 2007. Daily computer usage correlated with undergraduate students' musculoskeletal symptoms. Am. J. Ind. Med. 50, 481e488. Chowdhury, S.K., Nimbarte, A.D., Jaridi, M., Creese, R.C., 2013. Discrete wavelet transform analysis of surface electromyography for the fatigue assessment of neck and shoulder muscles. J. Electromyogr. Kinesiol. 23, 995e1003. ^ te , P., van der Velde, G., David, C.J., et al., 2008. The burden and determinants of Co neck pain in workers: results of the bone and joint decade 2000e2010 task force on neck pain and its associated disorders. Spine 33, S60eS74. Czerwinski, M., Smith, G., Regan, T., Meyers, B., et al., 2003. Toward Characterizing the Productivity Benefits of Very Large Displays. Human Computer Interaction, pp. 9e16. Dennerlein, J.T., Johnson, P.W., 2006. Different computer tasks affect the exposure of the upper extremity to biomechanical risk factors. Ergonomics 49, 45e55. Eltayeb, S., Staal, J.B., Hassan, A., De Bie, R.A., 2009. Work related risk factors for neck, shoulder and arms complaints: a cohort study among Dutch computer office workers. J. Occup. Rehabil. 19, 315e322.
Fitts, P.M., 1954. The information capacity of the human motor system in controlling the amplitude of movement. J. Exp. Psychol. 46, 381e391. Grandjean, E., Hunting, W., Nishiyama, K., 1984. Preferred VDT workstation settings, body posture and physical impairments. Appl. Ergon. 15, 99e104. Grudin, J., 2001. Partitioning digital worlds: focal and peripheral awareness in multiple monitor use. In: Proceedings of ACM CHI’92 Conference on Human Factors in Computing Systems, pp. 458e465. Hakala, P.T., Rimpela, A.H., Saarni, L.A., Salminen, J.J., 2006. Frequent computerrelated activities increase the risk of neck-shoulder and low back pain in adolescents. Eur. J. Public Health 16, 536e541. Hu, B., Ning, X., 2015. The changes of trunk motion rhythm and spinal loading during trunk flexion and extension motions caused by lumbar muscle fatigue. Ann. Biomed. Eng. 43, 2112e2119. Huang, H., Lai, H., 2008. Factors influencing the usability of icons in the LCD touchscreen. Sci. Direct 29, 339e344. Juul-Kristensen, B., Jensen, C., 2005. Self-reported workplace related ergonomic conditions as prognostic factors for musculoskeletal symptoms: the “BIT” follow up study on office workers. Occup. Environ. Med. 62, 188e194. Jones, M., Marsden, G., Mohd-Nasir, N., Boone, K., Buchanan, G., 1999. Improving web interaction on small displays. Comput. Netw. 31, 1129e1137. Jorgensen, K., Fallentin, N., Krogh-Lund, C., Jensen, B., 1988. Electromyography and fatigue during prolonged, low-level static contractions. Eur. J. Appl. Physiol. 57, 316e321. Kang, H., Shin, G., 2014. Hand usage pattern and upper body discomfort of desktop touchscreen users. Ergonomics 57. Karwowski, W., Eberts, R., Salvendy, G., Noland, S., 1994. The effects of computer interface design on human postural dynamics. Ergonomics 37, 703e724. Laughery, K.R., Young, S.L., Vaubel, K.P., Brelsfor, J.W., 1993. The noticeability of warnings on alcoholic beverage containers. J. Public Policy Mark. 12, 38e56. Lindsay, P.H., Norman, D.A., 1977. Human Information Processing: an Introduction to Psychology, second ed. Academic Press, London. Lutz, S., Huitt, W., 2003. Information Processing and Memory: Theory and Applications. In: Educational Psychology Interactive. Valdosta State University, Valdosta, GA. Retrievd [data], from. http://www.edpsycinteractive.org/papers/ infoproc.pdf. Martin, B.I., Deyo, R.A., Mirza, S.K., Turner, J.A., Comstock, B.A., Hollingworth, W., Sullivan, S.D., 2008. Expenditures and health status among adults with back and neck problems. JAMA 299, 656e664. Ming, Z., Narhi, M., Siivola, J., 2004. Neck and shoulder pain related to computer use. Pathophysiology 11, 51e56. Montgomery, D.C., 2012. Design and Analysis of Experiments, eighth ed. John Wiley and Sons, New York. Ning, X., Huang, Y., Hu, B., Nimbarte, A.D., 2015. Neck kinematics and muscle activity during mobile device operations. Int. J. Ind. Ergon. 48, 10e15. Nussbaum, M.A., 2001. Static and dynamic myoelectric measures of shoulder muscle fatigue during intermittent dynamic exertions of low to moderate
50
B. Hu, X. Ning / International Journal of Industrial Ergonomics 56 (2016) 41e50
intensity. Eur. J. Appl. Physiol. 85, 299e309. Oldfield, R.C., 1971. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 9, 97e113. Orvis, K.A., Horn, D.B., Belanich, J., 2008. The roles of task difficulty and prior videogame experience on performance and motivation in instructional videogames. Comput. Hum. Behav. 24, 2415e2433. Pincivero, D.M., Aldworth, C., Dickerson, T., Petry, C., Shultz, T., 2000. Quadriceps hamstring EMG activity during functional, closed kinetic chain exercise to fatigue. Eur. J. Appl. Physiol. 81, 504e509. Przybyla, A.S., Skrzypiec, D., Pollintine, P., Lolan, P., Adamas, M.A., 2007. Strength of the cervical spine in compression and bending. Spine 32, 1612e1620. SENIAM, 2012. Recommendations for Sensor Locations in Shoulder or Neck Muscles [online]. Available from: http://www.seniam.org. Schneider, W., Shiffrin, R.M., 1977. Controlled and automatic human information processing: I. Detection, search, and attention. Psychol. Rev. 84, 1e66. Shin, G., Zhu, X., 2011. User discomfort, work posture and muscle activity while using a touchscreen in a desktop PC setting. Ergonomics 54, 733e744. Sim, J., Lacey, R.J., Lewis, M., 2006. The impact of workplace risk factors on the occurrence of neck and upper limb pain: a general population study. BMC Public Health 6, 234.
Soukoreff, R.W., MacKenzie, I.S., 2004. Towards a standard for pointing device evaluation, perspectives on 27 years of Fitts' law research in HCI. Int. J. Hum. Comput. Stud. 61, 751e789. Staudenmann, D., Taube, W., 2015. Brachialis muscle activity can be assessed with surface electromyography. J. Electromyogr. Kinesiol. 25, 199e204. Woollacott, E., 2013. Using Windows 8 on Touchscreen Laptops Could Be a Pain in the Shoulder [press release on the internet]. Available from: http://www. theguardian.com/technology/2013/apr/09/windows-8-touchscreen-laptopspain. Young, J.G., Trudeau, M.B., Odell, D., Marinelli, K., Dennerlein, J.T., 2012. Touchscreen tablet user configurations and case-supported tilt affect head and neck flexion angles. Work 41, 81e91. Young, J.G., Trudeau, M.B., Odell, D., Marinelli, K., Dennerlein, J.T., 2013. Wrist and shoulder posture and muscle activity during touch-screen tablet use: effects of usage configuration, tablet type, and interacting hand. Work 45, 59e71. Zhou, J., Ning, X., Nimbarte, A.D., Dai, F., 2015. The assessment of material-handling strategies in dealing with sudden loading: the effect of uneven ground surface on trunk biomechanical response. Ergonomics 58, 259e267. Zhu, X., Shin, G., 2011. Shoulder and neck muscle activities during typing with articulating forearm support at different heights. Ergonomics 55, 1412e1419.