Applied Ergonomics 50 (2015) 98e104
Contents lists available at ScienceDirect
Applied Ergonomics journal homepage: www.elsevier.com/locate/apergo
Mobile input device type, texting style and screen size influence upper extremity and trapezius muscle activity, and cervical posture while texting David M. Kietrys a, Michael J. Gerg b, c, Jonathan Dropkin d, Judith E. Gold e, f, * a
School of Health Related Professions, Rutgers, The State University of New Jersey, 40 E. Laurel Road, Stratford, NJ, USA Occupational Therapy Assistant Program, Harcum College, 750 Montgomery Ave., Bryn Mawr, PA, USA Department of Rehabilitation Sciences, Temple University, 3307 N. Broad St., Philadelphia, PA 19140, USA d Department of Occupational Medicine, Epidemiology, and Prevention, Hofstra North Shore-LIJ School of Medicine, 175 Community Drive, Great Neck, NY 11021, USA e €vle, 801 76 Ga €vle, Sweden Centre for Musculoskeletal Research, University of Ga f Department of Public Health, Temple University, 3307 N. Broad St., Philadelphia, PA 19140, USA1 b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 3 July 2013 Accepted 2 March 2015 Available online
This study aimed to determine the effects of input device type, texting style, and screen size on upper extremity and trapezius muscle activity and cervical posture during a short texting task in college students. Users of a physical keypad produced greater thumb, finger flexor, and wrist extensor muscle activity than when texting with a touch screen device of similar dimensions. Texting on either device produced greater wrist extensor muscle activity when texting with 1 hand/thumb compared with both hands/thumbs. As touch screen size increased, more participants held the device on their lap, and chose to use both thumbs less. There was also a trend for greater finger flexor, wrist extensor, and trapezius muscle activity as touch screen size increased, and for greater cervical flexion, although mean differences for cervical flexion were small. Future research can help inform whether the ergonomic stressors observed during texting are associated with musculoskeletal disorder risk. © 2015 Elsevier Ltd and The Ergonomics Society. All rights reserved.
Keywords: Text messaging Ergonomic exposure SMS (short message service)
1. Introduction Globally, the number of mobile device subscriptions is approaching 6.8 billion (ITU, 2013). It is expected that worldwide text messaging frequency will increase to 9.4 trillion annually by 2016, up from 5.9 trillion in 2011 (Clark-Dickson, 2012). In the United States, it has been found that 91% of mobile device users perform text messaging, e-mailing, or instant messaging (Rainie, 2010). Individuals in a one Canadian university community spent more than 3.5 h per day using mobile devices to text, email, or browse the Internet (Berolo et al., 2011).
* Corresponding author. Present address: Centre for Musculoskeletal Research (CBF), Department of Occupational and Public Health Sciences, University of G€ avle, 801 76 G€ avle, Sweden. Tel.: þ46 (0)26 64 8624. E-mail addresses: [email protected] (D.M. Kietrys), mgerg@harcum. edu (M.J. Gerg), [email protected] (J. Dropkin), [email protected] (J.E. Gold). 1 Address at which authors performed experiment. http://dx.doi.org/10.1016/j.apergo.2015.03.003 0003-6870/© 2015 Elsevier Ltd and The Ergonomics Society. All rights reserved.
Case reports and epidemiologic studies indicate there is a potential risk of musculoskeletal symptoms and musculoskeletal disorders (MSDs) in the neck, shoulder, and hands with increased exposure to mobile device use for text messaging (Berolo et al., 2011; Gold et al., 2009; Korpinen et al., 2013; Menz, 2005; Ming et al., 2006; Storr et al., 2007; Williams and Kennedy, 2011). Nonneutral and static postures and forceful exertions have been identified as risk factors for neck/shoulder and hand/wrist MSDs (Bernard, 1997; National Academy of Sciences, 2001). Texting on mobile devices requires muscle activity to hold the device and type while texting. The moment generated through key activation (typically by the thumb) must be counterbalanced, likely through activation of the finger flexors and wrist extensors. Aspects of mobile device design, including input device type (touch screen or physical keypad), device weight, and screen size may influence ergonomic exposures while texting. Ergonomic exposures may be different when performing the same texting task on touch screen tablets of different sizes. Young et al. (2012) observed increased cervical flexion with touch screen placement
D.M. Kietrys et al. / Applied Ergonomics 50 (2015) 98e104
on one's lap vs. placement on a table. Due to their increased weight, mobile devices with larger touch screen sizes are more likely to be used on the lap as compared with smaller devices. Therefore, greater cervical flexion may occur with larger screen size. However, no previous studies have examined touch screen size and its influence on upper extremity muscle activity, upper trapezius muscle activity, and cervical posture. Texting styles (preferences for which digits and hands are used) may affect muscle activity, and joint angles. One observational study of texters (n ¼ 859) found the most frequent styles were both hands holding the device with both thumbs typing (46.1%) and one thumb typing while the same hand holds the device (44.1%) (Gold et al., 2012). Literature is sparse with regard to comparisons of muscle activities in the fingers, forearm/wrist, and neck and forearm/wrist and neck posture while texting in these two different styles. However, Gustafsson et al. (2011) found that greater extensor digitorum muscle activity was used during texting while holding the phone with one hand rather than two, which was attributed to higher stabilizing forces required when texting and gripping the phone with one hand. The wrist extensors work synergistically with the extensor digitorum (Kendall et al., 1993). Therefore, wrist extensor muscle activity would likely be associated with texting style. Mobile devices with the QWERTY keyboard configuration are generally of two input types: physical keypad and touch screen. Keystroke forces as measured through a force plate were greater in physical keypads than in touch screen keyboards (Kim et al., 2014). Similarly, participants in texting studies have been found to increase their applied muscle activity with increased computer keyboard key activation force requirements (Lee et al., 2009; Radwin and Ruffalo, 1999; Rempel et al., 1997. Therefore, users will likely exert greater thumb muscle forces when using a physical keypad than a touch screen mobile device. Individual keys on touch screen devices cannot be identified through haptic feedback. Typically, users must view these virtual keys to identify them. Generally, smaller keys are present on devices with smaller screen sizes. Thus, individuals may prefer to hold touch screen phones with smaller screens closer to their line of sight, using less cervical flexion to view the screen. To understand the potential effects of text messaging exposures on MSDs, it is important to describe muscle activity and joint angles while texting on commonly utilized hand held devices with commonly used texting styles. The study of muscles forces and postures associated with texting on hand held devices with different input types or different sizes, and texting with different styles could identify potentially negative musculoskeletal impacts on users and could inform preventive guidelines. Any adverse effect of input device type would be of particular import for mobile device designers, while an effect of typing style would be critical for users. In this study, we used readily available, commonly used, off-theshelf hand held devices. The primary aim of the study was to ascertain whether input device type and/or texting style affected upper extremity muscle activity and cervical posture during a short texting task. Because of the greater force required to activate physical keys, we hypothesize that greater thenar eminence activity will occur when using a physical keypad than a touch screen mobile device. Similarly, because the moment generated in texting must be counterbalanced by the finger flexors and wrist extensors, we hypothesize that wrist extensor and finger muscle activity will be also affected by input device type. Because of the increased need to stabilize and grip the phone with one-handed texting versus two-handed texting, we hypothesize that wrist extensor muscle activity will be influenced by texting style. We also sought to characterize participants' preferred texting styles with touch screen devices of different size
99
screens. Additionally, we explored the effects of texting on upper extremity and upper trapezius muscle activity and cervical posture with respect to touch screen size while holding the device with two hands and typing with both thumbs. 2. Material and methods 2.1. Participants College students were recruited via email and advertisements posted at a large urban university campus. Potential participants were screened over the phone for eligibility. Inclusion criteria included: at least 18 years of age, right hand dominant, and experience with use of mobile hand-held devices for text messaging. Exclusion criteria included allergies to adhesive tape, arm or hand pain or numbness rated 4 or higher on a 0e10 numerical scale, and experiencing arm or hand pain or numbness more than occasionally (i.e., frequently or continuously). Twenty eligible participants provided informed consent. This study was approved by the Temple University and University of Medicine & Dentistry of New Jersey Institutional Review Boards. 2.2. Texting procedures and conditions Participants sat in a wooden chair without armrests (Fig. 1) and texted the phrase “hi how are you” without capitalization or punctuation, at their preferred pace, and without stopping to correct errors. This task was executed continuously and repeatedly
Fig. 1. Subject texting in the one hand/one thumb configuration. Open circles represent spherical markers. Cervical flexion angle is formed between the dotted horizontal line and the dashed line connecting the forehead and C7. The black oval represents the surface EMG electrode over the extensor carpi radialis. EMG electrodes were also placed over the flexor digitorum superficialis, abductor pollicis brevis, and upper trapezius (not pictured).
100
D.M. Kietrys et al. / Applied Ergonomics 50 (2015) 98e104
over a period of 10 s. There was a break of 1 min between randomized experimental conditions. Two trials of each condition were executed. Participants were asked to rest for 30 s between trials of the same condition. The second trial only was used in data analysis. Prior to executing each condition, an investigator provided verbal instructions and demonstrated the condition. Mobile devices utilized in the study had a QWERTY format for character input. These included a physical keypad phone, 96 g, 8.4 4.3 2.3 cm, keyboard: 3.5 5.0 cm (LG 900G, LG Mobile Phones, San Diego, CA); a 3.500 touch screen phone, 101 g, 11.2 5.9 .71.0 cm, keyboard: 2.7 4.3 cm (iPod touch model MC544LL, Apple, Cupertino, CA); a 700 touch screen notepad, 385 g, 18.6 12.0 1.2 cm (Samsung Galaxy Tab model GT-P1010, Samsung Electronics, Ridgefield Park, NJ); and a 9.500 touch screen notepad, 603 g, 18.6 24.1 86.0 cm, (iPad 2 model MC770LL, Apple, Cupertino, CA). Each mobile device was made ready for texting before being handed to the participant. That is, “Notes” or another similar app on the mobile device was activated before the participant was given the device. Each of these applications had a font size of 8.5 points. The randomized conditions consisted of the following: physical keypad device, right hand holding device, right thumb typing (1/1) physical keypad device, both hands holding device, both thumbs typing (2/2) 3.500 touch screen device, right hand holding device, right thumb typing (1/1) 3.500 touch screen device, both hands holding device, both thumbs typing (2/2) 3.500 touch screen device: preferred style 700 touch screen device: preferred style 9.500 touch screen device: preferred style For the first four listed conditions, participants were required to hold the device in a portrait orientation. For the preferred style conditions, mobile devices were placed screen down on a table located 1 m away from the participant. Participants retrieved the device, resumed sitting in the chair, and spent as much time as desired to determine how they wanted to hold the device and which fingers they wanted to use to text. In general, participants self-determined how they wanted to hold the device in less than 1 min. After they began texting, a researcher (DK or MG) recorded which fingers were used and whether the device or its bottom edge was on the lap. 2.3. Outcome measures 2.3.1. Cervical posture To measure cervical posture in the sagittal plane, adhesive spherical markers (3/800 in diameter) were affixed to the participants over the 7th cervical vertebrae spinous process and the midforehead. A video camera (HDR-SR11, Sony Corporation; Tokyo, Japan) was affixed to a tripod located 1.5 m from the participant. Camera height was adjusted for each participant so that the center of the lens was at the same height as the participant's external auditory meatus while sitting in the chair. Prior to the texting trials, participants were asked to sit upright in a wooden chair and fix their gaze upon an eye-level wall graphic located 4.6 m away for a period of 10 s during which a lateral view digital video (30 Hz) was recorded. This was regarded as the cervical neutral condition. In data analysis, for the cervical neutral condition and each texting condition video, markers were digitized in and a line between the markers was drawn (Fig. 1). The angle between this line and the horizontal was calculated (MaxTRAQ V2.2, Innovision
Systems Inc., Columbiaville, MI). This angle was calculated for all frames of the video clip and the mean angle was utilized as the data point. The amount of forward motion (cervical flexion) for a given texting condition was determined by subtracting the mean texting condition cervical angle from the mean cervical neutral angle. 2.3.2. Muscle activity Bi-polar differential surface EMG electrodes with an interelectrode distance of 20 mm (SX320, Biometrics Ltd., Newport, UK) were affixed with double-stick tape to 4 muscles groups on the participant's right side: (1) upper trapezius, (2) extensor carpi radialis (longus and brevis) (ECR), (3) flexor digitorum superficialis (FDS), and (4) abductor pollicis brevis (APB) prior to the texting trials. The electrode for upper trapezius was placed at the midpoint between C7 and the acromion; the electrode for the extensor carpi radialis was placed 2 fingerbreadths distal to the lateral epicondyle; the electrode for the flexor digitorum superficialis was placed midway along the volar aspect of the forearm, slightly medial to midline; the electrode for the abductor pollicis brevis was placed midway between the volar aspect of the first metacarpalphalangeal joint and the first carpometacarpal joint (Delagi et al., 2011). A ground electrode was affixed to the ulnar styloid process of the left wrist. The electrodes were connected to the same DataLOG device and associated software described above, thus enabling synchronization with the electrogoniometer data for wrist motion. The EMG sampling rate was 1000 Hz. For all EMG measurements, the root mean square (RMS) signal was used with a moving average over 0.1 s. The 50th percentile RMS signal was used in data analysis. Prior to the texting trials, for each of the four muscle groups, the surface EMG signal was collected during three, 5-s maximum voluntary contractions (MVCs), separated by a 1 min break between contractions to allow for muscle recovery. A minimum break of at least 1 min was taken between testing of each muscle group. The first and last 0.25 s of each exertion was removed, in order to eliminate any transients. The highest value of the 3 exertions was designated as the 100% MVC. MVCs were obtained as follows: Upper trapezius: Participants were seated in a chair with their arms at their sides and asked to shrug upward while grasping metal handles affixed to the floor with a steel chain. ECR: Participants were seated (with their elbow bent 90 and forearm pronated and resting on a table) and grasped a metal handle affixed to the floor with a steel chain. FDS: Participants gripped a hydraulic hand dynamometer (Jamar model 5030J1, Lafayette Instruments, Lafayette, IN). APB: Participants performed a key-style pinch using a hydraulic pinch dynamometer (Baseline model 12-0200, Fabrication Enterprises, White Plains, NY). Surface EMG activity from each of the electrodes was recorded during each texting trial. 2.5. Statistical analysis Each muscle activity and joint angle variable was tested for normality with the ShapiroeWilk test and by viewing PP and QQ plots. Since the muscle activity variables were not normally distributed, these were log transformed in the data analysis. Mean muscle activities and joint angles were computed on a per condition basis. A repeated measures ANOVA (RMANOVA) was performed to evaluate the effect of input device type, texting style and interaction between input device type and texting style on the mean muscle activities and cervical posture between the 4 conditions identified for our primary aim (physical keypad device and touch screen device, each with 1/1 or 2/2 texting styles).
D.M. Kietrys et al. / Applied Ergonomics 50 (2015) 98e104
The 2/2 texting style was selected as a preferred style by some participants for all three screen sizes. A repeated measures ANOVA was performed to evaluate the effect of screen size on mean muscle activity and cervical posture while participants typed in the 2/2 style. Since there were three levels of screen size, Mauchly's test of sphericity was conducted. If sphericity could not be assumed, the Greenhouse-Geisser adjustment for the p-value was used. Trends in outcome variables by screen size were determined using a test for linear trend in SPSS. Data are presented as mean and standard deviation. Statistical significance was regarded at p < 0.05. Excel (Microsoft Corp., Redmond, WA), SAS V9.3 (SAS Institute Inc., Cary, NC) and SPSS V20 (IBM Corp., Armonk, NY) were used in data analysis.
3. Results 3.1. Participants
101
Table 2 Preferred texting styles on touch screen mobile devices, by screen size (n ¼ 20). Screen size
3.500 700 9.500 a
Texting style, n (%) Both thumbs only
Right index finger, with any combination of other digits
Device on lapa
18 (90%) 15 (75%) 5 (25%)
0 (0%) 3 (15%) 14 (70%)
0 (0%) 2 (10%) 12 (60%)
Either the entire device or the bottom edge of device resting on lap.
3.2.2. Cervical flexion For cervical flexion there was an effect of texting style (RMANOVA, p ¼
Contents lists available at ScienceDirect
Applied Ergonomics journal homepage: www.elsevier.com/locate/apergo
Mobile input device type, texting style and screen size influence upper extremity and trapezius muscle activity, and cervical posture while texting David M. Kietrys a, Michael J. Gerg b, c, Jonathan Dropkin d, Judith E. Gold e, f, * a
School of Health Related Professions, Rutgers, The State University of New Jersey, 40 E. Laurel Road, Stratford, NJ, USA Occupational Therapy Assistant Program, Harcum College, 750 Montgomery Ave., Bryn Mawr, PA, USA Department of Rehabilitation Sciences, Temple University, 3307 N. Broad St., Philadelphia, PA 19140, USA d Department of Occupational Medicine, Epidemiology, and Prevention, Hofstra North Shore-LIJ School of Medicine, 175 Community Drive, Great Neck, NY 11021, USA e €vle, 801 76 Ga €vle, Sweden Centre for Musculoskeletal Research, University of Ga f Department of Public Health, Temple University, 3307 N. Broad St., Philadelphia, PA 19140, USA1 b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 3 July 2013 Accepted 2 March 2015 Available online
This study aimed to determine the effects of input device type, texting style, and screen size on upper extremity and trapezius muscle activity and cervical posture during a short texting task in college students. Users of a physical keypad produced greater thumb, finger flexor, and wrist extensor muscle activity than when texting with a touch screen device of similar dimensions. Texting on either device produced greater wrist extensor muscle activity when texting with 1 hand/thumb compared with both hands/thumbs. As touch screen size increased, more participants held the device on their lap, and chose to use both thumbs less. There was also a trend for greater finger flexor, wrist extensor, and trapezius muscle activity as touch screen size increased, and for greater cervical flexion, although mean differences for cervical flexion were small. Future research can help inform whether the ergonomic stressors observed during texting are associated with musculoskeletal disorder risk. © 2015 Elsevier Ltd and The Ergonomics Society. All rights reserved.
Keywords: Text messaging Ergonomic exposure SMS (short message service)
1. Introduction Globally, the number of mobile device subscriptions is approaching 6.8 billion (ITU, 2013). It is expected that worldwide text messaging frequency will increase to 9.4 trillion annually by 2016, up from 5.9 trillion in 2011 (Clark-Dickson, 2012). In the United States, it has been found that 91% of mobile device users perform text messaging, e-mailing, or instant messaging (Rainie, 2010). Individuals in a one Canadian university community spent more than 3.5 h per day using mobile devices to text, email, or browse the Internet (Berolo et al., 2011).
* Corresponding author. Present address: Centre for Musculoskeletal Research (CBF), Department of Occupational and Public Health Sciences, University of G€ avle, 801 76 G€ avle, Sweden. Tel.: þ46 (0)26 64 8624. E-mail addresses: [email protected] (D.M. Kietrys), mgerg@harcum. edu (M.J. Gerg), [email protected] (J. Dropkin), [email protected] (J.E. Gold). 1 Address at which authors performed experiment. http://dx.doi.org/10.1016/j.apergo.2015.03.003 0003-6870/© 2015 Elsevier Ltd and The Ergonomics Society. All rights reserved.
Case reports and epidemiologic studies indicate there is a potential risk of musculoskeletal symptoms and musculoskeletal disorders (MSDs) in the neck, shoulder, and hands with increased exposure to mobile device use for text messaging (Berolo et al., 2011; Gold et al., 2009; Korpinen et al., 2013; Menz, 2005; Ming et al., 2006; Storr et al., 2007; Williams and Kennedy, 2011). Nonneutral and static postures and forceful exertions have been identified as risk factors for neck/shoulder and hand/wrist MSDs (Bernard, 1997; National Academy of Sciences, 2001). Texting on mobile devices requires muscle activity to hold the device and type while texting. The moment generated through key activation (typically by the thumb) must be counterbalanced, likely through activation of the finger flexors and wrist extensors. Aspects of mobile device design, including input device type (touch screen or physical keypad), device weight, and screen size may influence ergonomic exposures while texting. Ergonomic exposures may be different when performing the same texting task on touch screen tablets of different sizes. Young et al. (2012) observed increased cervical flexion with touch screen placement
D.M. Kietrys et al. / Applied Ergonomics 50 (2015) 98e104
on one's lap vs. placement on a table. Due to their increased weight, mobile devices with larger touch screen sizes are more likely to be used on the lap as compared with smaller devices. Therefore, greater cervical flexion may occur with larger screen size. However, no previous studies have examined touch screen size and its influence on upper extremity muscle activity, upper trapezius muscle activity, and cervical posture. Texting styles (preferences for which digits and hands are used) may affect muscle activity, and joint angles. One observational study of texters (n ¼ 859) found the most frequent styles were both hands holding the device with both thumbs typing (46.1%) and one thumb typing while the same hand holds the device (44.1%) (Gold et al., 2012). Literature is sparse with regard to comparisons of muscle activities in the fingers, forearm/wrist, and neck and forearm/wrist and neck posture while texting in these two different styles. However, Gustafsson et al. (2011) found that greater extensor digitorum muscle activity was used during texting while holding the phone with one hand rather than two, which was attributed to higher stabilizing forces required when texting and gripping the phone with one hand. The wrist extensors work synergistically with the extensor digitorum (Kendall et al., 1993). Therefore, wrist extensor muscle activity would likely be associated with texting style. Mobile devices with the QWERTY keyboard configuration are generally of two input types: physical keypad and touch screen. Keystroke forces as measured through a force plate were greater in physical keypads than in touch screen keyboards (Kim et al., 2014). Similarly, participants in texting studies have been found to increase their applied muscle activity with increased computer keyboard key activation force requirements (Lee et al., 2009; Radwin and Ruffalo, 1999; Rempel et al., 1997. Therefore, users will likely exert greater thumb muscle forces when using a physical keypad than a touch screen mobile device. Individual keys on touch screen devices cannot be identified through haptic feedback. Typically, users must view these virtual keys to identify them. Generally, smaller keys are present on devices with smaller screen sizes. Thus, individuals may prefer to hold touch screen phones with smaller screens closer to their line of sight, using less cervical flexion to view the screen. To understand the potential effects of text messaging exposures on MSDs, it is important to describe muscle activity and joint angles while texting on commonly utilized hand held devices with commonly used texting styles. The study of muscles forces and postures associated with texting on hand held devices with different input types or different sizes, and texting with different styles could identify potentially negative musculoskeletal impacts on users and could inform preventive guidelines. Any adverse effect of input device type would be of particular import for mobile device designers, while an effect of typing style would be critical for users. In this study, we used readily available, commonly used, off-theshelf hand held devices. The primary aim of the study was to ascertain whether input device type and/or texting style affected upper extremity muscle activity and cervical posture during a short texting task. Because of the greater force required to activate physical keys, we hypothesize that greater thenar eminence activity will occur when using a physical keypad than a touch screen mobile device. Similarly, because the moment generated in texting must be counterbalanced by the finger flexors and wrist extensors, we hypothesize that wrist extensor and finger muscle activity will be also affected by input device type. Because of the increased need to stabilize and grip the phone with one-handed texting versus two-handed texting, we hypothesize that wrist extensor muscle activity will be influenced by texting style. We also sought to characterize participants' preferred texting styles with touch screen devices of different size
99
screens. Additionally, we explored the effects of texting on upper extremity and upper trapezius muscle activity and cervical posture with respect to touch screen size while holding the device with two hands and typing with both thumbs. 2. Material and methods 2.1. Participants College students were recruited via email and advertisements posted at a large urban university campus. Potential participants were screened over the phone for eligibility. Inclusion criteria included: at least 18 years of age, right hand dominant, and experience with use of mobile hand-held devices for text messaging. Exclusion criteria included allergies to adhesive tape, arm or hand pain or numbness rated 4 or higher on a 0e10 numerical scale, and experiencing arm or hand pain or numbness more than occasionally (i.e., frequently or continuously). Twenty eligible participants provided informed consent. This study was approved by the Temple University and University of Medicine & Dentistry of New Jersey Institutional Review Boards. 2.2. Texting procedures and conditions Participants sat in a wooden chair without armrests (Fig. 1) and texted the phrase “hi how are you” without capitalization or punctuation, at their preferred pace, and without stopping to correct errors. This task was executed continuously and repeatedly
Fig. 1. Subject texting in the one hand/one thumb configuration. Open circles represent spherical markers. Cervical flexion angle is formed between the dotted horizontal line and the dashed line connecting the forehead and C7. The black oval represents the surface EMG electrode over the extensor carpi radialis. EMG electrodes were also placed over the flexor digitorum superficialis, abductor pollicis brevis, and upper trapezius (not pictured).
100
D.M. Kietrys et al. / Applied Ergonomics 50 (2015) 98e104
over a period of 10 s. There was a break of 1 min between randomized experimental conditions. Two trials of each condition were executed. Participants were asked to rest for 30 s between trials of the same condition. The second trial only was used in data analysis. Prior to executing each condition, an investigator provided verbal instructions and demonstrated the condition. Mobile devices utilized in the study had a QWERTY format for character input. These included a physical keypad phone, 96 g, 8.4 4.3 2.3 cm, keyboard: 3.5 5.0 cm (LG 900G, LG Mobile Phones, San Diego, CA); a 3.500 touch screen phone, 101 g, 11.2 5.9 .71.0 cm, keyboard: 2.7 4.3 cm (iPod touch model MC544LL, Apple, Cupertino, CA); a 700 touch screen notepad, 385 g, 18.6 12.0 1.2 cm (Samsung Galaxy Tab model GT-P1010, Samsung Electronics, Ridgefield Park, NJ); and a 9.500 touch screen notepad, 603 g, 18.6 24.1 86.0 cm, (iPad 2 model MC770LL, Apple, Cupertino, CA). Each mobile device was made ready for texting before being handed to the participant. That is, “Notes” or another similar app on the mobile device was activated before the participant was given the device. Each of these applications had a font size of 8.5 points. The randomized conditions consisted of the following: physical keypad device, right hand holding device, right thumb typing (1/1) physical keypad device, both hands holding device, both thumbs typing (2/2) 3.500 touch screen device, right hand holding device, right thumb typing (1/1) 3.500 touch screen device, both hands holding device, both thumbs typing (2/2) 3.500 touch screen device: preferred style 700 touch screen device: preferred style 9.500 touch screen device: preferred style For the first four listed conditions, participants were required to hold the device in a portrait orientation. For the preferred style conditions, mobile devices were placed screen down on a table located 1 m away from the participant. Participants retrieved the device, resumed sitting in the chair, and spent as much time as desired to determine how they wanted to hold the device and which fingers they wanted to use to text. In general, participants self-determined how they wanted to hold the device in less than 1 min. After they began texting, a researcher (DK or MG) recorded which fingers were used and whether the device or its bottom edge was on the lap. 2.3. Outcome measures 2.3.1. Cervical posture To measure cervical posture in the sagittal plane, adhesive spherical markers (3/800 in diameter) were affixed to the participants over the 7th cervical vertebrae spinous process and the midforehead. A video camera (HDR-SR11, Sony Corporation; Tokyo, Japan) was affixed to a tripod located 1.5 m from the participant. Camera height was adjusted for each participant so that the center of the lens was at the same height as the participant's external auditory meatus while sitting in the chair. Prior to the texting trials, participants were asked to sit upright in a wooden chair and fix their gaze upon an eye-level wall graphic located 4.6 m away for a period of 10 s during which a lateral view digital video (30 Hz) was recorded. This was regarded as the cervical neutral condition. In data analysis, for the cervical neutral condition and each texting condition video, markers were digitized in and a line between the markers was drawn (Fig. 1). The angle between this line and the horizontal was calculated (MaxTRAQ V2.2, Innovision
Systems Inc., Columbiaville, MI). This angle was calculated for all frames of the video clip and the mean angle was utilized as the data point. The amount of forward motion (cervical flexion) for a given texting condition was determined by subtracting the mean texting condition cervical angle from the mean cervical neutral angle. 2.3.2. Muscle activity Bi-polar differential surface EMG electrodes with an interelectrode distance of 20 mm (SX320, Biometrics Ltd., Newport, UK) were affixed with double-stick tape to 4 muscles groups on the participant's right side: (1) upper trapezius, (2) extensor carpi radialis (longus and brevis) (ECR), (3) flexor digitorum superficialis (FDS), and (4) abductor pollicis brevis (APB) prior to the texting trials. The electrode for upper trapezius was placed at the midpoint between C7 and the acromion; the electrode for the extensor carpi radialis was placed 2 fingerbreadths distal to the lateral epicondyle; the electrode for the flexor digitorum superficialis was placed midway along the volar aspect of the forearm, slightly medial to midline; the electrode for the abductor pollicis brevis was placed midway between the volar aspect of the first metacarpalphalangeal joint and the first carpometacarpal joint (Delagi et al., 2011). A ground electrode was affixed to the ulnar styloid process of the left wrist. The electrodes were connected to the same DataLOG device and associated software described above, thus enabling synchronization with the electrogoniometer data for wrist motion. The EMG sampling rate was 1000 Hz. For all EMG measurements, the root mean square (RMS) signal was used with a moving average over 0.1 s. The 50th percentile RMS signal was used in data analysis. Prior to the texting trials, for each of the four muscle groups, the surface EMG signal was collected during three, 5-s maximum voluntary contractions (MVCs), separated by a 1 min break between contractions to allow for muscle recovery. A minimum break of at least 1 min was taken between testing of each muscle group. The first and last 0.25 s of each exertion was removed, in order to eliminate any transients. The highest value of the 3 exertions was designated as the 100% MVC. MVCs were obtained as follows: Upper trapezius: Participants were seated in a chair with their arms at their sides and asked to shrug upward while grasping metal handles affixed to the floor with a steel chain. ECR: Participants were seated (with their elbow bent 90 and forearm pronated and resting on a table) and grasped a metal handle affixed to the floor with a steel chain. FDS: Participants gripped a hydraulic hand dynamometer (Jamar model 5030J1, Lafayette Instruments, Lafayette, IN). APB: Participants performed a key-style pinch using a hydraulic pinch dynamometer (Baseline model 12-0200, Fabrication Enterprises, White Plains, NY). Surface EMG activity from each of the electrodes was recorded during each texting trial. 2.5. Statistical analysis Each muscle activity and joint angle variable was tested for normality with the ShapiroeWilk test and by viewing PP and QQ plots. Since the muscle activity variables were not normally distributed, these were log transformed in the data analysis. Mean muscle activities and joint angles were computed on a per condition basis. A repeated measures ANOVA (RMANOVA) was performed to evaluate the effect of input device type, texting style and interaction between input device type and texting style on the mean muscle activities and cervical posture between the 4 conditions identified for our primary aim (physical keypad device and touch screen device, each with 1/1 or 2/2 texting styles).
D.M. Kietrys et al. / Applied Ergonomics 50 (2015) 98e104
The 2/2 texting style was selected as a preferred style by some participants for all three screen sizes. A repeated measures ANOVA was performed to evaluate the effect of screen size on mean muscle activity and cervical posture while participants typed in the 2/2 style. Since there were three levels of screen size, Mauchly's test of sphericity was conducted. If sphericity could not be assumed, the Greenhouse-Geisser adjustment for the p-value was used. Trends in outcome variables by screen size were determined using a test for linear trend in SPSS. Data are presented as mean and standard deviation. Statistical significance was regarded at p < 0.05. Excel (Microsoft Corp., Redmond, WA), SAS V9.3 (SAS Institute Inc., Cary, NC) and SPSS V20 (IBM Corp., Armonk, NY) were used in data analysis.
3. Results 3.1. Participants
101
Table 2 Preferred texting styles on touch screen mobile devices, by screen size (n ¼ 20). Screen size
3.500 700 9.500 a
Texting style, n (%) Both thumbs only
Right index finger, with any combination of other digits
Device on lapa
18 (90%) 15 (75%) 5 (25%)
0 (0%) 3 (15%) 14 (70%)
0 (0%) 2 (10%) 12 (60%)
Either the entire device or the bottom edge of device resting on lap.
3.2.2. Cervical flexion For cervical flexion there was an effect of texting style (RMANOVA, p ¼
