- Email: [email protected]
Effect of computer mouse gain and visual demand on mouse clicking performance and muscle activation in a young and elderly group of experienced computer users
ARTICLE IN PRESS
Applied Ergonomics 36 (2005) 547–555 www.elsevier.com/locate/apergo
Effect of computer mouse gain and visual demand on mouse clicking performance and muscle activation in a young and elderly group of experienced computer users J. Sandfelda,, B.R. Jensenb a
Institute of Exercise and Sport Sciences, University of Copenhagen, Denmark Institute of Exercise and Sport Sciences, University of Copenhagen, Denmark, Nørre Alle 51, 2200-N Copenhagen, Denmark
b
Accepted 4 March 2005
Abstract The present study evaluated the specific effects of motor demand and visual demands on the ability to control motor output in terms of performance and muscle activation. Young and elderly subjects performed multidirectional pointing tasks with the computer mouse. Three levels of mouse gain and three levels of target size were used. All subjects demonstrated a reduced working speed and hit rate at the highest mouse gain (1:8) when the target size was small. The young group had an optimum at mouse gain 1:4. The elderly group was most sensitive to the combination of high mouse gain and small targets and thus, this age group should avoid this combination. Decreasing target sizes (i.e. increasing visual demand) reduced performance in both groups despite that motor demand was maintained constant. Therefore, it is recommended to avoid small screen objects and letters. Forearm muscle activity was only to a minor degree influenced by mouse gain (and target sizes) indicating that stability of the forearm/hand is of significance during computer mouse control. The study has implications for ergonomists, pointing device manufacturers and software developers. r 2005 Elsevier Ltd. All rights reserved. Keywords: Aging; Computer mouse; Motor control
1. Introduction At present, not only the young but also an increasing part of the elderly population uses computers and computer mice both at work and at home. Computer mouse work requires extensive hand-eye coordination skills and this fact combined with the well documented age related decline in fine motor control skills (Cooke et al., 1989; Ranganathan et al., 2001) is the background for a reduced performance during computer mouse clicking found in elderly (Laursen et al., 2001; Smith et al., 1999). Especially, when the demand for precision is increased the performance of the elderly is reduced. Therefore, the gain of the pointing device (e.g. the Corresponding author. Tel.: +45 3532 7220; fax: +45 3532 7217.
E-mail address: sandfeld@ifi.ku.dkdk (J. Sandfeld). 0003-6870/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.apergo.2005.03.003
computer mouse) may also play a significant role for the degree of motor control required to perform computer mouse tasks. If the computer mouse gain is high, even small movements of the device have great impact on the amplitude of the screen cursor movements. Thus, adjustments of the mouse gain serve as a potential tool for optimization of performance and musculoskeletal workload in relation to age. In general, the physical demands during computer mouse tasks involve motor demands as well as visual demands. The level of the motor demands is dependent on the movement amplitude in combination with the demands for precision of the hand movements whereas the visual demands are determined by e.g. the size of the targets and their inter-distance on the monitor. These individual components of the physical demands have not been studied separately. Systematic combinations of computer mouse
ARTICLE IN PRESS 548
J. Sandfeld, B.R. Jensen / Applied Ergonomics 36 (2005) 547–555
gain, sizes of screen targets and distances between targets make it possible to evaluate the specific effects of motor demands and visual demands on the ability to control motor output in terms of performance and muscle activation. Such results may help pointing device manufacturers, software developers, ergonomists and managers to optimize the conditions for the human when working at the computer. It is hypothesized that performance will decrease with increasing motor demand even when the visual demand is kept constant. Furthermore, we anticipate that performance at different levels of motor demands will influence young and elderly differently due to an expected reduction in motor control in the elderly group. Secondly, it is hypothesized that increasing visual demands will not reduce performance if the index of difficulty according to Fitts Principle is kept constant and if the smallest screen objects are visually easy to detect. The aim was to study the effect of various motor demands and visual demands on performance and muscle activation during computer mouse work in a young and an elderly group.
2. Methods 2.1. Participants A total of 33 experienced computer users participated, 17 young (26.572.6 years. 9 females, 8 males) and 16 elderly (64.973.6 years. 8 females, 8 males). The young group had 6.372.8 years of experience using the computer mouse and the average computer use per week exceeded 20 h. The elderly had 7.374.0 years of experience and the average computer use per week exceeded 18 h. The Nordic questionnaire (Kuorinka et al., 1987) and additional questions regarding present health status were used to assure that only persons who to their own knowledge had no muscular or neurological disorders or symptoms were included in the study. The study was approved by the Local Danish Committee of Ethics and the participants gave their informed written consent before participation. 2.2. Visual acuity Participants were only included if they to their own knowledge had normal or corrected to normal vision by lenses, reading or bifocal/progressive glasses. Furthermore, a modified Snellen vision acuity test was used to compare the visual capabilities of the two groups at normal screen distance. Seven lines with letters of decreasing font size from eleven to five point was shown one at the time on the computer screen. Each line consisted of seven letters. Letters of font size five
represented the smallest possible font on the computer screen, because the distance between the lines in the letters was one pixel. The average distance from eye to screen of all the participants was 62711 cm (SD) and all were able to read the presented font sizes. 2.3. Experimental setup The participants were seated in an adjustable chair and at a height-adjustable table (Ergoscandia, Slagelse, Denmark) which was used to allow for complete elbow and forearm support for both arms when working with the computer mouse. Starting position was with the right upper arm placed vertical next to the truncus, abducted approximately 101 and with the forearm in horizontal position pointing forward. A digitizer (Wacom Intous GD0912P PC, tablet size A3, sensitivity 0.25 mm) with mouse was used for registering computer mouse position and clicking. In addition, a standard PC (Hewlett Packard) with custom made software was used along with a high resolution (17 in, 1024 768 pixels) Cathode Ray Tube (CRT) screen (Sony ES500). The screen was placed in the personal preferred position and the average distance from the eyes of the participants to the screen was 65711 cm for the young group and 58710 cm for the elderly group. The top of the screen was 5.474.8 cm below eye level in the young group and 4.973.5 cm in the elderly group. 2.4. Experimental design The subjects performed three sessions of a multidirectional pointing task (International Organization for Standardization, 1997) (ISO 9241-9) where the object was to click on successively appearing square icons (targets). Upon clicking a target a new target would appear on the opposite site of an invisible circle 1981 from the preceding position so that clicking would be performed in all directions slowly moving clockwise. A total of 20 clicks completed one full circle (Laursen et al., 2001). The three clicking sessions were performed with a computer mouse gain (MG) of 1:2, 1:4 and 1:8, respectively (ratios denote mouse movement relative to screen cursor movement and is equivalent to mouse speed without acceleration in most software programs). Each session included two conditions (described below) with three 70-s trials in each condition. The three trials were performed using three different target sizes, each appearing in a circle with a diameter, which according to Fitts’ principle resulted in a constant index of difficulty. Target sizes were: small (8 8 pixels, circle diameter 128 pixels), medium (16 16 pixels wide, circle diameter 256 pixels) or large (32 32 pixels, circle diameter 512 pixels) (1 pixel ¼ 0.315 mm). Index of difficulty (Id) was calculated using the ISO-standard 9241: I d ¼ Log2 (ðD þ W Þ=W ), where D is the distance between targets
ARTICLE IN PRESS J. Sandfeld, B.R. Jensen / Applied Ergonomics 36 (2005) 547–555
and W is the width of the targets. In the present study I d was kept constant at 4.09 for the three target sizes. According to the ISO standard I d between 4 and 6 represents a medium accuracy demand required for a pointing task. The size of the smallest targets equalled the size of 12–14 font size letters on a 1700 CRT monitor and thus, represented a common target size when working with text editing, spreadsheets or graphics. The medium and large targets represented common size icons or buttons in toolbars or on web pages. In the first condition (fastest possible clicking frequency (FastCF)) the participants were instructed to ‘‘click from one target to the next as fast as possibly but try to avoid making errors’’ (error ¼ not hitting a target). In the second working condition (Constant clicking frequency (ConCF)) the clicking frequency was constant at 50 Hz guided by a metronome. This allowed between group comparison of performance and muscle activity. The participants were instructed to ‘‘click on the targets but primarily, click on every beat from the metronome’’. Thus, a click was required even if the screen cursor was outside the screen target. The order of the three sessions (MG 1:2, 1:4 & 1:8) was randomized
549
and the order of trials (target sizes) within the first session was randomized but was the same for FastCF and ConCF and across all three sessions. The participants practiced at each MG for 3 min before each session. Preliminary testing showed that performance stabilized after 3 min practice with a new mouse gain. Between each of the 70-s trials, there was a resting period of approximately 30 s and 3 min rest was allowed between sessions before changing to a new MG. The design allowed the effect of motor demands and visual demands during work with the computer mouse to be studied separately. Thus, the effect of varied motor demands while the visual input from the computer screen remained unchanged was evaluated by extracting data from the three MG at the two different conditions. This was done for the small, the medium and the large target sizes (totally 9 trials 2 conditions) (Fig. 1A). The influence of altering the visual demands while motor demands, i.e. mouse movements, remained unchanged was measured by extracting results from the combination of large targets at MG 1:8, medium targets at MG 1:4 and small targets at MG 1:2. This was also done for both conditions (6 trials). In all six trials the demanded
Visual demand - Large target Motor demand - MG 1:2 - MG 1:4 - MG 1:8
(A)
0
20
40
80
160
0
20
40
80
160
Visual demand - Large target
- Medium target - Small target Motor demand
(B)
Distance between targets or amplitude of movement (mm)
Fig. 1. (A) An example of altering motor demand at the large target size. Increasing the mouse gain (MG) from 1:2 to 1:8 increased motor demand while the visual demand was constant. This was also done for the medium and small target size. (B) Reducing target size from large to small while reducing MG from 1:8 to 1:2 decreased visual input while the motor demand was kept constant. White squares: visual demand, grey squares: motor demand.
ARTICLE IN PRESS 550
J. Sandfeld, B.R. Jensen / Applied Ergonomics 36 (2005) 547–555
computer mouse movement was 20.1 mm from target to target and the target width equalled a mouse movement of 1.25 mm (Fig. 1B).
reported from the ConCF condition where the working speed was the same for the two groups. 2.7. Statistics
2.5. Electromyography (EMG) The effect of changing MG and target size was evaluated for each age group using a two-way ANOVA (general linear model) (3 MG 3 target sizes). In case of FastCF Working speed (clicks/min)
75
*
1:4 1:8
*
65 60 55
*
50 45 40 Large
Medium
Small
Large
Young
(A)
Medium
Small
Elderly
FastCF 100 95
*
90 85 80 75 70 Large
Medium
Small
Large
Young
(B)
Medium
Small
Elderly
ConCF *
100
2.6. Data analysis
*
*
95 Hit rate (%)
Clicking frequency and errors were registered in the software controlling the multi-directional pointing task. The performance parameters, working speed (clicks/min) and relative hit rate representing the percentage of successful clicks were calculated based on the last 60 s of each 70-s trial. This was done to avoid bias from the start-up period of the trial. Working speed and hit rate were calculated during FastCF. Hit rate was calculated during ConCF. Isometric muscle strength and EMGmax were determined as the maximal mean value developed during 1 s and the highest value found in the three trials was used. Root mean square EMG (RMS-EMG) using intervals of 100 ms was then calculated and expressed relative to maximum EMG (%EMGmax). The RMS-EMG is
1:2
*
70
Hit rate (%)
Surface EMG electrodes were placed on m. extensor carpi radialis (ECR), m. extensor digitorum superficialis (ED), m. extensor carpi ulnaris (ECU), m. flexor carpi radialis (FCR)), right and left mm. trapezius (RTRAP, LTRAP) and right neck extensor (NE) muscle (upper part of m. trapezius and m. splenius capitis). The bipolar Ag/AgCl surface electrodes (720 01 K, Medicotest, Olstykke, Denmark) were placed on the most bulky part of each muscle with a center-to-center electrode distance of 20 mm. The EMG was recorded at 1024 Hz and lowpass filtered at 450 Hz (Butterworth filter) and highpass filtered at 10 Hz. For the EMG normalization, isometric maximal voluntary contractions (MVC) were performed during handgrip contraction, wrist extension and flexion, finger extension, ulnar deviation, shoulder elevation and neck extension. Handgrip was performed with the hand in neutral (thumb up) position with the forearm horizontal and a 901 flexion in the elbow. Wrist extension and flexion, finger extension and ulnar deviation were performed against firm supports with the forearm and hand supported in horizontal position with the forearm pronated. The forearm was fixated with a strap just proximal to the wrist. Shoulder elevation was performed with the subjects standing upright with an adjustable strap on each shoulder. Finally, the subjects performed maximum neck extension against his/her hands, which were folded behind the head (Laursen et al., 2001, 2002) . The MVCs were performed three times each. The EMG results are reported as relative to highest EMG value found during the three trials (%EMGmax).
90
*
85 80 75 70 Large
(C)
Medium Young
Small
Large
Medium
Small
Elderly
Fig. 2. The effect on the two age groups when changing movement precision and range (mouse gain) at each of the three target sizes (large, medium and small). (A) Working speed (clicks/min) and (B) hit rate (% successful hits) during FastCF (the condition where participants were asked to click as fast as possible). (C) Hit rate during ConCF (clicking frequency was guided by a metronome at 50 clicks/min). *: po0:05.
ARTICLE IN PRESS J. Sandfeld, B.R. Jensen / Applied Ergonomics 36 (2005) 547–555
FastCF Large target / 1:8 75 #
Working speed (Clicks/min)
interaction, univariate ANOVA with a Tukey post hoc test was used for each age group across target sizes to compare differences between MG. The two age groups were compared by collapsing the target sizes and using a two-way ANOVA with repeated measurements (2 age groups 3 MG). Differences in performance between the three levels of motor demand while keeping the visual demand constant were detected using a multiple ANOVA (MANOVA). Same procedure was used to detect difference in performance when changing visual demand while keeping motor demand constant. Data are presented as means7standard deviation in the text and as means7standard error in the figures. For level of significance pp0:05 was used.
60 *
55 50
Young
Elderly
FastCF 100
3.1. Effect of motor demand on performance
* *
Hit rate(%)
95
90
85
80 (B)
Young
Elderly
ConCF
100
* *
Hit rate (%)
95
90
85
80 (C)
Fig. 3A shows that there was a tendency to a reduced working speed (p ¼ 0:56) in the young group with increasing visual demand. The corresponding result was significant in the elderly group where the working speed was reduced by 10.4%. When the visual demand was increased from large to small target size the hit rate during FastCF decreased in both age groups by 5.5%
Small target / 1:2
65
45
3. Results
3.2. Effect of visual demand on performance
Medium target / 1:4
70
(A)
Fig. 2A shows that the working speed was affected by the motor demand in the young group mainly expressed as a higher working speed at MG 1:4 at the large and medium target size. At the small target size a lower working speed of 13% and 9% in the young and elderly group, respectively, was found at MG 1:8 compared to the average working speed of MG 1:2 and 1:4. The level of motor demand did not influence the working speed of the elderly group at the larger target sizes. In general, the hit rate during FastCF was not affected by the motor demand in any of the age groups, except for a lower hit rate in the elderly group at the smallest target size (Fig. 2B). During the ConCF, at 50 clicks/min, the overall finding was that the hit rate was affected by the motor demand in both age groups (Fig. 2C). Interaction between MG and target size showed that the effect of motor demand increased in both age groups with decreasing target size. Hence, the MANOVA showed that hit rate was independent of the motor demand at the largest target size in both age groups. However, there was a major decrease in the hit rate with increasing motor demand at the smallest target, especially in the elderly group where the hit rate declined to 75% (Fig. 2C).
551
Young
Elderly
Fig. 3. The effect on the two age groups when changing visual demand and keeping movement precision and range constant. This was attainable when combining the data from mouse gain (MG) 1:8 with the large target, MG 1:4 with the medium target and MG 1:2 with the small target. (A) Working speed (clicks/min) and (B) hit rate (% successful hits) during FastCF (the condition where participants were asked to click as fast as possible). (C) Hit rate during ConCF (clicking frequency was guided by a metronome at 50 clicks/min). *: po0:05, #: po0:1.
ARTICLE IN PRESS J. Sandfeld, B.R. Jensen / Applied Ergonomics 36 (2005) 547–555
3.3. Effect on the two age-groups During FastCF, the elderly had an overall lower working speed of 22% compared to the young while there was no difference in hit rate between the young and elderly at any of the target sizes. During ConCF the elderly had a lower hit rate compared to the young. This was the case for all target sizes where the elderly had a 6.6% lower hit rate average compared to the young. Additionally, the elderly were more affected by the combination of increased motor demand and visual demand compared to the young. Hence, the hit rate of the elderly group decreased by 20.6% from the highest hit rate at MG 1:4 large target to the lowest at MG 1:8 small target. The corresponding value was 12.4% for the young group. 3.4. Muscle activity Average RMS-EMG during mouse clicking is reported from ConCF where the working speed was the same in both groups (Fig. 4). In the young group, when motor demand was increased, we found a higher RMSEMG across target sizes in ECR (7.7%), ED (7.4%) and FCR (41.4%) and a reduction in RTRAP (57.9%) and the NE (18%) (Fig. 4A). This means that the muscle activity level increased in three out of four forearm muscles even though the movement of the mouse decreased with increasing motor demand. In the elderly group an increase across target sizes was similarly found in ED (6.7%) and a reduction was found in RTRAP (49.1%) and NE (14.2%) with increasing motor demand (Fig. 4B). When increasing visual demand at a constant motor demand, the RMS-EMG in the young group showed a small increase in RTRAP and in the elderly group there was an increase in the FCR and a tendency to an increase in RTRAP. The grand average RMS-EMG was higher in the elderly compared to the young (Fig. 4). For the ECR the grand average RMS-EMG were 3.4% and 5.2%EMGmax in the young and the elderly group, respectively and for ED the values were 7.6%EMGmax for the young and 11.0%EMGmax for the elderly. For the ECU RMS-EMG were 7.8%EMGmax in the young group and 10.7%EMGmax in the elderly group and for the FCR, values were 1.2% and 2.8%EMGmax for the young and the elderly, respectively. RMS-EMG in RTRAP and LTRAP in the young group was 2.4% and 1.0%EMGmax, respectively whereas the values were 5.9% and 3.4%EMGmax in the elderly group.
Young 14.0 RMS-EMG (%EMGmax)
and 5.7% in the young and elderly group, respectively (Fig. 3B). Similarly, the hit rate decreased during ConCF by 4.8% in the young group and 7.6% in the elderly group (Fig. 3C).
1:2 1:4 1:8
12.0 10.0
*
8.0 6.0 4.0
* *
* *
2.0 0.0 (A)
ECR
ED
ECU
FCR
RTRAP LTRAP
NE
Elderly 14.0 RMS-EMG (%EMGmax)
552
*
*
12.0 10.0
*
8.0 6.0 4.0 2.0 0.0
(B)
ECR
ED
ECU
FCR RTRAP LTRAP
NE
Fig. 4. Muscle activity measured with electromyography (EMG). The effect of increasing mouse gain across three target sizes. ECR: m. extensor carpi radialis, ED: m. extensor digitorum superficialis, ECU: m. extensor carpi ulnaris, FCR: m. flexor carpi radialis, RTRAP and LTRAP: right and left m. trapezius, NE: right neck extensor muscle (upper part of m. trapezius and m. splenius capitis). (A) The young group, (B) the elderly group. EMG was reported from the session where working speed and thereby working load was kept constant by guidance of a metronome (50 clicks/min).
Finally, the values for NE were 4.1% and 9.9%EMGmax for the young and elderly, respectively.
4. Discussion The present study demonstrated a rather complex relationship between computer MG and performance. Thus, the young subjects seem to have an optimum at MG 1:4 provided that the target sizes were large or medium. This optimum was not found among the elderly. Additionally, at the highest MG a decrease in working speed and hit rate was found for both age groups, especially when the target size was small. Thus, our hypothesis regarding the relation between performance and motor demand was confirmed for the small targets but the interaction between the two parameters turned out to be more complex than originally suggested
ARTICLE IN PRESS J. Sandfeld, B.R. Jensen / Applied Ergonomics 36 (2005) 547–555
and hence, the hypothesis was not confirmed for the two largest targets. Still, the important notion can be made, that reducing MG may enhance performance when working with small screen objects. Because the MG determines the distance the computer mouse must be moved between targets, it may be expected that MG is related to the EMG-level. In both groups, the ED muscle activity and in the young group, the ECR and FCR muscle activity increased with increasing MG. However, the absolute differences in muscle activity across the three MG were small and probably without major functional relevance. This indicates that the activation of the forearm muscles may be determined to a higher degree by the demand for stability of the forearm/hand rather than by the movement of the hand per se. Furthermore, the combination of a high MG and small targets had a higher impact on performance in the elderly group compared to the young group. This may be explained by several factors. Working with the computer mouse utilizes mainly closed loop movement cycles, where movement causes afferent feed-back information which is processed in the brain and integrated with new motor programs, resulting in new movements. Possible factors involved in the reduced hand-eye control generally found in elderly are several. Morphological age-related changes such as death of neurons innervating the large type II muscle fibers, cause a reorganization of the motor unit composition resulting in enlargement of type I motor units (Doherty et al., 1993; Lexell, 1995). Larger type I motor units make it more difficult to graduate the force output at low intensity work like mouse handling. Age-related changes in neurological factors such as decline in the number of neurons (Vandervoort, 2002), reduction in motor unit excitability (Sabbahi and Sedgwich, 1982) and slowing of nerve conduction velocity (Dorfman and Bosley, 1979) also contribute to the decrease in motor control in elderly even though the impact of each of these factors is very difficult to quantify. Finally, slower information processing and motor programming in the brain which has been found in the elderly may be an influential factor, since a longer duration of each closed loop cycle reduces the number of corrections possible within a fixed time frame. Consequently, the execution of a task will either be less precise or take longer time for an elderly compared to a young person (Cerella, 1985; Rubichi et al., 1999; Van der Lubbe and Verleger, 2002). The generally higher EMG-level in the elderly compared to the young found in the present study has also been found in earlier studies using the multidirectional pointing task and it is speculated whether this phenomenon may be the result of a higher central drive and/or a higher antagonist activity in the elderly. It has been shown that overall higher EMG-level increase joint stiffness which stabilize the movement and reduce
553
the acceleration of movements (Milner and Cloutier, 1998; Seidler-Dobrin et al., 1998). Thus, in contrast to the above mentioned factors the higher EMG-level in the elderly may be a compensatory mechanism which in part counteracts the decline in motor control. As the reported RMS-EMG is the percentage of the maximal EMG, it could be speculated if the higher relative EMGlevel in the elderly is a result of a reduced ability of the elderly to activate the muscle maximally during the MVC. However, several studies addressing this topic show that the active elderly are able to fully activate the available muscle mass (Connelly et al., 1999; De Serres and Enoka, 1998; Jakobi and Rice, 2002; Klein et al., 2001; Phillips et al., 1992). In spite of between group differences in muscle activity level, similar activation pattern between muscles were found in the two age groups. Finally, we found that concentrated mouse work demands a relatively high level of muscle activity in especially ED and ECU, but also in the NE in the elderly group. When considering the often long duration of every day mouse work, this may be a factor contributing to development of work related muscular skeletal symptoms. According to our second hypothesis we did not expect to find a relationship between visual demand and performance. However, the present results indicate that increased visual demand reduced performance even though index of difficulty was kept constant. This may be explained by the limitations of the human visual acuity, which among other things depends on the area of the retina stimulated. With decreasing target size the area of the retina stimulated was decreased and hence, a reduction of the ratio between the visual signal and the uncertainty in the visual system occurred. That is, there was a reduction in the signal-to-noise ratio which caused a decline in quality of the visual object projected in the visual cortex. However, the average person with normal vision can resolve points separated by one minute of arc (¼ 1=60 of one degree) in a traditional Snellen test but the size of the smallest screen target (2.5 2.5 mm) in the present study equalled approximately 14 min of arc and should thus, in theory not constitute any visual difficulties (Davson, 1990). However, the task was to place the cursor within the target and thus the subjects had to visually detect space between the cursor tip and the edge of the target, which substantially increases the visual demand. Furthermore, when working at the computer screen and performing a multi-directional pointing task, visual acuity is influenced by additional factors which potentially increases the demand on the visual system compared to the Snellen test. For instance, in contrast to the Snellen test where the subject is situated 6 m from the chart, the image at the screen is located at a near or intermediate distance of approximately 60 cm which means that accommodation and vergence may play a role in maintaining clarity of the
ARTICLE IN PRESS 554
J. Sandfeld, B.R. Jensen / Applied Ergonomics 36 (2005) 547–555
image (Kurimoto et al., 1983). The time factor also influences the performance because time is needed for binocular refixation whenever a new target appeared (Adrian, 2003; Kaufman et al., 1992). Finally, factors such as the degree of illumination and contrast which has a great influence on visual acuity may have been different in the present task compared to the settings needed for testing with the Snellen chart. Taken together, the computer task may have constituted a higher visual demand than what could be expected based merely on the screen target sizes. Earlier studies of motor control related to computer work have found a reduction in performance in the elderly compared to the young and specifically, it has been shown that even though elderly have the same hit rate during a multi-directional pointing task compared to young, this outcome is attained on the expense of the working speed which is lower in the elderly. However, when a demand for a certain clicking frequency was applied to the clicking task, a reduced hit rate has been found in the elderly compared to the young. Also, it has been found that increasing precision demand reduces the performance of the elderly to a larger extent than the young (Laursen et al., 2001; Smith et al., 1999). The present findings confirmed these earlier results as the working speed was lower in the elderly group compared to the young, while no difference in hit rate was found. Fitts’ index of difficulty have been widely used in the human-computer interaction literature and recommended for comparison of usability of non-keyboard input devices. The index of difficulty is based on the assumption that target width and inter-target distances are related in a manner based on information technology (Fitts, 1954). Fitts’ index of difficulty was kept constant in all trials in the present study and therefore, one should expect a similar level of performance in all the trials. However, the data showed major variations in the performance, depending on the combination of mouse gain and target size. One likely explanation for these findings is that Fitts’ index of difficulty was originally validated in a one-to-one setting where the amplitude of the arm/hand movement was directly related to the visual input, which is rare during use of computer input devices. Based on this, it is recommended that usability evaluations and comparisons of input devices using the ISO-standard, which is an evolution of the Fitts equation should include the gain of the pointing device as an additional parameter that alters task difficulty. Otherwise wrong conclusions regarding usability of the pointing device may be drawn.
5. Conclusion The study contributes to the understanding of the interaction between the physical demands involved in
use of computer mouse, performance and muscle activation in relation to aging. In both age groups performance was reduced with increasing visual demands, indicating that computer mouse work with small screen objects and letters should be avoided if possible. The combination of small target sizes and high mouse gain reduced performance severely and this was especially pronounced in the elderly group. Thus, if working with small screen objects or letters, it is recommended to decrease mouse gain. A mouse gain of 1:4 seems to be optimal for young computer users. The muscle activation levels were in general higher for the elderly than the young group, although the activation strategy was similar. Finally, it is recommended to expand the guidelines for testing input device usability, to include pointing device gain as task difficulty is dependent on input device gain.
Acknowledgements We would like to thank the Danish Research Agency for supporting the project. Also we would like to acknowledge the technical assistance of Marianne Pilegaard, Institute of Exercise and Sport Sciences and Bjarne Laursen, National Institute of Public Health. References Adrian, W., 2003. The effect of observation time and contrast on visual acuity. Clin. Exp. Optometry 86 (3), 179–182. Cerella, J., 1985. Information-processing rates in the elderly. Psychol. Bull. 98 (1), 67–83. Connelly, D.M., Rice, C.L., Roos, M.R., Vandervoort, A.A., 1999. Motor unit firing rates and contractile properties in tibialis anterior of young and old men. J. Appl. Physiol. 87 (2), 843–852. Cooke, J.D., Brown, S.H., Cunningham, D.A., 1989. Kinematics of arm movements in elderly humans. Neurobiol. Aging 10, 159–165. Davson, H., 1990. Physiology of the Eye. Macmillan Press, London. De Serres, S.J., Enoka, R.M., 1998. Older adults can maximally activate the biceps brachii muscle by voluntary command. Am. Physiol. Soc. 84 (1), 284–291. Doherty, T.J., Vandervoort, A.A., Taylor, A.W., Brown, W.F., 1993. Effects of motor unit losses on strength in older men and women. J. Appl. Physiol. 74 (2), 868–874. Dorfman, L.J., Bosley, T.M., 1979. Age-related changes in peripheral and central nerve conduction in man. Neurology 29, 38–44. Fitts, P.M., 1954. The information capacity of the human motor system in controlling the amplitude of movement. J. Exp. Psychol. 47 (6), 381–391. International Organization for Standardization, 1997. ISO 9241, Part 9 CD: Non-keyboard Input Devices, 1–47. Jakobi, J.M., Rice, C.L., 2002. Voluntary muscle activation varies with age and muscle group. J. Appl. Physiol. 93 (2), 457–462. Kaufman, P.L., Adler, F.H., Hart, W.M., Moses, R.A., 1992. Adler’s physiology of the eye. Mosby Year Book, St. Louis. Klein, C.S., Rice, C.L., Marsh, G.D., 2001. Normalized force, activation, and coactivation in the arm muscles of young and old men. J. Appl. Physiol. 91 (3), 1341–1349.
ARTICLE IN PRESS J. Sandfeld, B.R. Jensen / Applied Ergonomics 36 (2005) 547–555 Kuorinka, I., Jonsson, B., Kilbom, A˚., Vinterberg, H., BieringSørensen, F., Andersson, G., Jørgensen, K., 1987. Standardised Nordic questionnaires for the analysis of musculoskeletal symptoms. Appl. Ergonomics 18 (3), 233–237. Kurimoto, S., Iwasaki, T., Nomura, T., Noro, K., Yamamoto, S., 1983. Influence of VDT (visual display terminals) work on eye accommodation. J.UOEH 5 (1), 101–110. Laursen, B., Jensen, B.R., Ratkevicius, A., 2001. Performance and muscle activity during computer mouse tasks in young and elderly adults. Eur. J. Appl. Physiol. Occup. Physiol. 84 (4), 329–336. Laursen, B., Jensen, B.R., Garde, A.H., Jørgensen, A.H., 2002. Effect of mental and physical demands on muscular activity during the use of a computer mouse and a keyboard. Scand. J. Work Environ. Health 28 (4), 215–221. Lexell, J., 1995. Human aging, muscle mass, and fiber type composition. J. Gerontol. 50A (Special Issue), 11–16. Milner, T.E., Cloutier, C., 1998. Damping of the wrist joint during voluntary movement. Exp. Brain Res. 122 (3), 309–317.
555
Phillips, S.K., Bruce, S.A., Newton, D., Woledge, R.C., 1992. The weakness of old age is not due to failure of muscle activation. J. Gerontol. 47 (2), M45–M49. Ranganathan, V.K., Siemionow, V., Sahgal, V., Yue, G.H., 2001. Effects of aging on hand function. J. Am. Geriatr. Soc. 49 (11), 1478–1484. Rubichi, S., Neri, M., Nicoletti, R., 1999. Age-related slowing of control processes: evidence from a response coordination task. Cortex 35 (4), 573–582. Sabbahi, M.A., Sedgwich, E.M., 1982. Age-related changes in monosynaptic reflex excitability. J. Gerontol. 37 (1), 24–32. Seidler-Dobrin, R.D., He, J., Stelmach, G.E., 1998. Coactivation to reduce variability in the elderly. Motor Control 2, 314–330. Smith, M.W., Sharit, J., Czaja, S.J., 1999. Aging, motor control, and the performance of computer mouse tasks. Human Factors 41 (3), 389–396. Van der Lubbe, R.H.J., Verleger, R., 2002. Aging and the Simon task. Psychophysiology 39 (1), 100–110. Vandervoort, A.A., 2002. Aging of the human neuromuscular system. Muscle Nerve 25 (1), 17–25.
Applied Ergonomics 36 (2005) 547–555 www.elsevier.com/locate/apergo
Effect of computer mouse gain and visual demand on mouse clicking performance and muscle activation in a young and elderly group of experienced computer users J. Sandfelda,, B.R. Jensenb a
Institute of Exercise and Sport Sciences, University of Copenhagen, Denmark Institute of Exercise and Sport Sciences, University of Copenhagen, Denmark, Nørre Alle 51, 2200-N Copenhagen, Denmark
b
Accepted 4 March 2005
Abstract The present study evaluated the specific effects of motor demand and visual demands on the ability to control motor output in terms of performance and muscle activation. Young and elderly subjects performed multidirectional pointing tasks with the computer mouse. Three levels of mouse gain and three levels of target size were used. All subjects demonstrated a reduced working speed and hit rate at the highest mouse gain (1:8) when the target size was small. The young group had an optimum at mouse gain 1:4. The elderly group was most sensitive to the combination of high mouse gain and small targets and thus, this age group should avoid this combination. Decreasing target sizes (i.e. increasing visual demand) reduced performance in both groups despite that motor demand was maintained constant. Therefore, it is recommended to avoid small screen objects and letters. Forearm muscle activity was only to a minor degree influenced by mouse gain (and target sizes) indicating that stability of the forearm/hand is of significance during computer mouse control. The study has implications for ergonomists, pointing device manufacturers and software developers. r 2005 Elsevier Ltd. All rights reserved. Keywords: Aging; Computer mouse; Motor control
1. Introduction At present, not only the young but also an increasing part of the elderly population uses computers and computer mice both at work and at home. Computer mouse work requires extensive hand-eye coordination skills and this fact combined with the well documented age related decline in fine motor control skills (Cooke et al., 1989; Ranganathan et al., 2001) is the background for a reduced performance during computer mouse clicking found in elderly (Laursen et al., 2001; Smith et al., 1999). Especially, when the demand for precision is increased the performance of the elderly is reduced. Therefore, the gain of the pointing device (e.g. the Corresponding author. Tel.: +45 3532 7220; fax: +45 3532 7217.
E-mail address: sandfeld@ifi.ku.dkdk (J. Sandfeld). 0003-6870/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.apergo.2005.03.003
computer mouse) may also play a significant role for the degree of motor control required to perform computer mouse tasks. If the computer mouse gain is high, even small movements of the device have great impact on the amplitude of the screen cursor movements. Thus, adjustments of the mouse gain serve as a potential tool for optimization of performance and musculoskeletal workload in relation to age. In general, the physical demands during computer mouse tasks involve motor demands as well as visual demands. The level of the motor demands is dependent on the movement amplitude in combination with the demands for precision of the hand movements whereas the visual demands are determined by e.g. the size of the targets and their inter-distance on the monitor. These individual components of the physical demands have not been studied separately. Systematic combinations of computer mouse
ARTICLE IN PRESS 548
J. Sandfeld, B.R. Jensen / Applied Ergonomics 36 (2005) 547–555
gain, sizes of screen targets and distances between targets make it possible to evaluate the specific effects of motor demands and visual demands on the ability to control motor output in terms of performance and muscle activation. Such results may help pointing device manufacturers, software developers, ergonomists and managers to optimize the conditions for the human when working at the computer. It is hypothesized that performance will decrease with increasing motor demand even when the visual demand is kept constant. Furthermore, we anticipate that performance at different levels of motor demands will influence young and elderly differently due to an expected reduction in motor control in the elderly group. Secondly, it is hypothesized that increasing visual demands will not reduce performance if the index of difficulty according to Fitts Principle is kept constant and if the smallest screen objects are visually easy to detect. The aim was to study the effect of various motor demands and visual demands on performance and muscle activation during computer mouse work in a young and an elderly group.
2. Methods 2.1. Participants A total of 33 experienced computer users participated, 17 young (26.572.6 years. 9 females, 8 males) and 16 elderly (64.973.6 years. 8 females, 8 males). The young group had 6.372.8 years of experience using the computer mouse and the average computer use per week exceeded 20 h. The elderly had 7.374.0 years of experience and the average computer use per week exceeded 18 h. The Nordic questionnaire (Kuorinka et al., 1987) and additional questions regarding present health status were used to assure that only persons who to their own knowledge had no muscular or neurological disorders or symptoms were included in the study. The study was approved by the Local Danish Committee of Ethics and the participants gave their informed written consent before participation. 2.2. Visual acuity Participants were only included if they to their own knowledge had normal or corrected to normal vision by lenses, reading or bifocal/progressive glasses. Furthermore, a modified Snellen vision acuity test was used to compare the visual capabilities of the two groups at normal screen distance. Seven lines with letters of decreasing font size from eleven to five point was shown one at the time on the computer screen. Each line consisted of seven letters. Letters of font size five
represented the smallest possible font on the computer screen, because the distance between the lines in the letters was one pixel. The average distance from eye to screen of all the participants was 62711 cm (SD) and all were able to read the presented font sizes. 2.3. Experimental setup The participants were seated in an adjustable chair and at a height-adjustable table (Ergoscandia, Slagelse, Denmark) which was used to allow for complete elbow and forearm support for both arms when working with the computer mouse. Starting position was with the right upper arm placed vertical next to the truncus, abducted approximately 101 and with the forearm in horizontal position pointing forward. A digitizer (Wacom Intous GD0912P PC, tablet size A3, sensitivity 0.25 mm) with mouse was used for registering computer mouse position and clicking. In addition, a standard PC (Hewlett Packard) with custom made software was used along with a high resolution (17 in, 1024 768 pixels) Cathode Ray Tube (CRT) screen (Sony ES500). The screen was placed in the personal preferred position and the average distance from the eyes of the participants to the screen was 65711 cm for the young group and 58710 cm for the elderly group. The top of the screen was 5.474.8 cm below eye level in the young group and 4.973.5 cm in the elderly group. 2.4. Experimental design The subjects performed three sessions of a multidirectional pointing task (International Organization for Standardization, 1997) (ISO 9241-9) where the object was to click on successively appearing square icons (targets). Upon clicking a target a new target would appear on the opposite site of an invisible circle 1981 from the preceding position so that clicking would be performed in all directions slowly moving clockwise. A total of 20 clicks completed one full circle (Laursen et al., 2001). The three clicking sessions were performed with a computer mouse gain (MG) of 1:2, 1:4 and 1:8, respectively (ratios denote mouse movement relative to screen cursor movement and is equivalent to mouse speed without acceleration in most software programs). Each session included two conditions (described below) with three 70-s trials in each condition. The three trials were performed using three different target sizes, each appearing in a circle with a diameter, which according to Fitts’ principle resulted in a constant index of difficulty. Target sizes were: small (8 8 pixels, circle diameter 128 pixels), medium (16 16 pixels wide, circle diameter 256 pixels) or large (32 32 pixels, circle diameter 512 pixels) (1 pixel ¼ 0.315 mm). Index of difficulty (Id) was calculated using the ISO-standard 9241: I d ¼ Log2 (ðD þ W Þ=W ), where D is the distance between targets
ARTICLE IN PRESS J. Sandfeld, B.R. Jensen / Applied Ergonomics 36 (2005) 547–555
and W is the width of the targets. In the present study I d was kept constant at 4.09 for the three target sizes. According to the ISO standard I d between 4 and 6 represents a medium accuracy demand required for a pointing task. The size of the smallest targets equalled the size of 12–14 font size letters on a 1700 CRT monitor and thus, represented a common target size when working with text editing, spreadsheets or graphics. The medium and large targets represented common size icons or buttons in toolbars or on web pages. In the first condition (fastest possible clicking frequency (FastCF)) the participants were instructed to ‘‘click from one target to the next as fast as possibly but try to avoid making errors’’ (error ¼ not hitting a target). In the second working condition (Constant clicking frequency (ConCF)) the clicking frequency was constant at 50 Hz guided by a metronome. This allowed between group comparison of performance and muscle activity. The participants were instructed to ‘‘click on the targets but primarily, click on every beat from the metronome’’. Thus, a click was required even if the screen cursor was outside the screen target. The order of the three sessions (MG 1:2, 1:4 & 1:8) was randomized
549
and the order of trials (target sizes) within the first session was randomized but was the same for FastCF and ConCF and across all three sessions. The participants practiced at each MG for 3 min before each session. Preliminary testing showed that performance stabilized after 3 min practice with a new mouse gain. Between each of the 70-s trials, there was a resting period of approximately 30 s and 3 min rest was allowed between sessions before changing to a new MG. The design allowed the effect of motor demands and visual demands during work with the computer mouse to be studied separately. Thus, the effect of varied motor demands while the visual input from the computer screen remained unchanged was evaluated by extracting data from the three MG at the two different conditions. This was done for the small, the medium and the large target sizes (totally 9 trials 2 conditions) (Fig. 1A). The influence of altering the visual demands while motor demands, i.e. mouse movements, remained unchanged was measured by extracting results from the combination of large targets at MG 1:8, medium targets at MG 1:4 and small targets at MG 1:2. This was also done for both conditions (6 trials). In all six trials the demanded
Visual demand - Large target Motor demand - MG 1:2 - MG 1:4 - MG 1:8
(A)
0
20
40
80
160
0
20
40
80
160
Visual demand - Large target
- Medium target - Small target Motor demand
(B)
Distance between targets or amplitude of movement (mm)
Fig. 1. (A) An example of altering motor demand at the large target size. Increasing the mouse gain (MG) from 1:2 to 1:8 increased motor demand while the visual demand was constant. This was also done for the medium and small target size. (B) Reducing target size from large to small while reducing MG from 1:8 to 1:2 decreased visual input while the motor demand was kept constant. White squares: visual demand, grey squares: motor demand.
ARTICLE IN PRESS 550
J. Sandfeld, B.R. Jensen / Applied Ergonomics 36 (2005) 547–555
computer mouse movement was 20.1 mm from target to target and the target width equalled a mouse movement of 1.25 mm (Fig. 1B).
reported from the ConCF condition where the working speed was the same for the two groups. 2.7. Statistics
2.5. Electromyography (EMG) The effect of changing MG and target size was evaluated for each age group using a two-way ANOVA (general linear model) (3 MG 3 target sizes). In case of FastCF Working speed (clicks/min)
75
*
1:4 1:8
*
65 60 55
*
50 45 40 Large
Medium
Small
Large
Young
(A)
Medium
Small
Elderly
FastCF 100 95
*
90 85 80 75 70 Large
Medium
Small
Large
Young
(B)
Medium
Small
Elderly
ConCF *
100
2.6. Data analysis
*
*
95 Hit rate (%)
Clicking frequency and errors were registered in the software controlling the multi-directional pointing task. The performance parameters, working speed (clicks/min) and relative hit rate representing the percentage of successful clicks were calculated based on the last 60 s of each 70-s trial. This was done to avoid bias from the start-up period of the trial. Working speed and hit rate were calculated during FastCF. Hit rate was calculated during ConCF. Isometric muscle strength and EMGmax were determined as the maximal mean value developed during 1 s and the highest value found in the three trials was used. Root mean square EMG (RMS-EMG) using intervals of 100 ms was then calculated and expressed relative to maximum EMG (%EMGmax). The RMS-EMG is
1:2
*
70
Hit rate (%)
Surface EMG electrodes were placed on m. extensor carpi radialis (ECR), m. extensor digitorum superficialis (ED), m. extensor carpi ulnaris (ECU), m. flexor carpi radialis (FCR)), right and left mm. trapezius (RTRAP, LTRAP) and right neck extensor (NE) muscle (upper part of m. trapezius and m. splenius capitis). The bipolar Ag/AgCl surface electrodes (720 01 K, Medicotest, Olstykke, Denmark) were placed on the most bulky part of each muscle with a center-to-center electrode distance of 20 mm. The EMG was recorded at 1024 Hz and lowpass filtered at 450 Hz (Butterworth filter) and highpass filtered at 10 Hz. For the EMG normalization, isometric maximal voluntary contractions (MVC) were performed during handgrip contraction, wrist extension and flexion, finger extension, ulnar deviation, shoulder elevation and neck extension. Handgrip was performed with the hand in neutral (thumb up) position with the forearm horizontal and a 901 flexion in the elbow. Wrist extension and flexion, finger extension and ulnar deviation were performed against firm supports with the forearm and hand supported in horizontal position with the forearm pronated. The forearm was fixated with a strap just proximal to the wrist. Shoulder elevation was performed with the subjects standing upright with an adjustable strap on each shoulder. Finally, the subjects performed maximum neck extension against his/her hands, which were folded behind the head (Laursen et al., 2001, 2002) . The MVCs were performed three times each. The EMG results are reported as relative to highest EMG value found during the three trials (%EMGmax).
90
*
85 80 75 70 Large
(C)
Medium Young
Small
Large
Medium
Small
Elderly
Fig. 2. The effect on the two age groups when changing movement precision and range (mouse gain) at each of the three target sizes (large, medium and small). (A) Working speed (clicks/min) and (B) hit rate (% successful hits) during FastCF (the condition where participants were asked to click as fast as possible). (C) Hit rate during ConCF (clicking frequency was guided by a metronome at 50 clicks/min). *: po0:05.
ARTICLE IN PRESS J. Sandfeld, B.R. Jensen / Applied Ergonomics 36 (2005) 547–555
FastCF Large target / 1:8 75 #
Working speed (Clicks/min)
interaction, univariate ANOVA with a Tukey post hoc test was used for each age group across target sizes to compare differences between MG. The two age groups were compared by collapsing the target sizes and using a two-way ANOVA with repeated measurements (2 age groups 3 MG). Differences in performance between the three levels of motor demand while keeping the visual demand constant were detected using a multiple ANOVA (MANOVA). Same procedure was used to detect difference in performance when changing visual demand while keeping motor demand constant. Data are presented as means7standard deviation in the text and as means7standard error in the figures. For level of significance pp0:05 was used.
60 *
55 50
Young
Elderly
FastCF 100
3.1. Effect of motor demand on performance
* *
Hit rate(%)
95
90
85
80 (B)
Young
Elderly
ConCF
100
* *
Hit rate (%)
95
90
85
80 (C)
Fig. 3A shows that there was a tendency to a reduced working speed (p ¼ 0:56) in the young group with increasing visual demand. The corresponding result was significant in the elderly group where the working speed was reduced by 10.4%. When the visual demand was increased from large to small target size the hit rate during FastCF decreased in both age groups by 5.5%
Small target / 1:2
65
45
3. Results
3.2. Effect of visual demand on performance
Medium target / 1:4
70
(A)
Fig. 2A shows that the working speed was affected by the motor demand in the young group mainly expressed as a higher working speed at MG 1:4 at the large and medium target size. At the small target size a lower working speed of 13% and 9% in the young and elderly group, respectively, was found at MG 1:8 compared to the average working speed of MG 1:2 and 1:4. The level of motor demand did not influence the working speed of the elderly group at the larger target sizes. In general, the hit rate during FastCF was not affected by the motor demand in any of the age groups, except for a lower hit rate in the elderly group at the smallest target size (Fig. 2B). During the ConCF, at 50 clicks/min, the overall finding was that the hit rate was affected by the motor demand in both age groups (Fig. 2C). Interaction between MG and target size showed that the effect of motor demand increased in both age groups with decreasing target size. Hence, the MANOVA showed that hit rate was independent of the motor demand at the largest target size in both age groups. However, there was a major decrease in the hit rate with increasing motor demand at the smallest target, especially in the elderly group where the hit rate declined to 75% (Fig. 2C).
551
Young
Elderly
Fig. 3. The effect on the two age groups when changing visual demand and keeping movement precision and range constant. This was attainable when combining the data from mouse gain (MG) 1:8 with the large target, MG 1:4 with the medium target and MG 1:2 with the small target. (A) Working speed (clicks/min) and (B) hit rate (% successful hits) during FastCF (the condition where participants were asked to click as fast as possible). (C) Hit rate during ConCF (clicking frequency was guided by a metronome at 50 clicks/min). *: po0:05, #: po0:1.
ARTICLE IN PRESS J. Sandfeld, B.R. Jensen / Applied Ergonomics 36 (2005) 547–555
3.3. Effect on the two age-groups During FastCF, the elderly had an overall lower working speed of 22% compared to the young while there was no difference in hit rate between the young and elderly at any of the target sizes. During ConCF the elderly had a lower hit rate compared to the young. This was the case for all target sizes where the elderly had a 6.6% lower hit rate average compared to the young. Additionally, the elderly were more affected by the combination of increased motor demand and visual demand compared to the young. Hence, the hit rate of the elderly group decreased by 20.6% from the highest hit rate at MG 1:4 large target to the lowest at MG 1:8 small target. The corresponding value was 12.4% for the young group. 3.4. Muscle activity Average RMS-EMG during mouse clicking is reported from ConCF where the working speed was the same in both groups (Fig. 4). In the young group, when motor demand was increased, we found a higher RMSEMG across target sizes in ECR (7.7%), ED (7.4%) and FCR (41.4%) and a reduction in RTRAP (57.9%) and the NE (18%) (Fig. 4A). This means that the muscle activity level increased in three out of four forearm muscles even though the movement of the mouse decreased with increasing motor demand. In the elderly group an increase across target sizes was similarly found in ED (6.7%) and a reduction was found in RTRAP (49.1%) and NE (14.2%) with increasing motor demand (Fig. 4B). When increasing visual demand at a constant motor demand, the RMS-EMG in the young group showed a small increase in RTRAP and in the elderly group there was an increase in the FCR and a tendency to an increase in RTRAP. The grand average RMS-EMG was higher in the elderly compared to the young (Fig. 4). For the ECR the grand average RMS-EMG were 3.4% and 5.2%EMGmax in the young and the elderly group, respectively and for ED the values were 7.6%EMGmax for the young and 11.0%EMGmax for the elderly. For the ECU RMS-EMG were 7.8%EMGmax in the young group and 10.7%EMGmax in the elderly group and for the FCR, values were 1.2% and 2.8%EMGmax for the young and the elderly, respectively. RMS-EMG in RTRAP and LTRAP in the young group was 2.4% and 1.0%EMGmax, respectively whereas the values were 5.9% and 3.4%EMGmax in the elderly group.
Young 14.0 RMS-EMG (%EMGmax)
and 5.7% in the young and elderly group, respectively (Fig. 3B). Similarly, the hit rate decreased during ConCF by 4.8% in the young group and 7.6% in the elderly group (Fig. 3C).
1:2 1:4 1:8
12.0 10.0
*
8.0 6.0 4.0
* *
* *
2.0 0.0 (A)
ECR
ED
ECU
FCR
RTRAP LTRAP
NE
Elderly 14.0 RMS-EMG (%EMGmax)
552
*
*
12.0 10.0
*
8.0 6.0 4.0 2.0 0.0
(B)
ECR
ED
ECU
FCR RTRAP LTRAP
NE
Fig. 4. Muscle activity measured with electromyography (EMG). The effect of increasing mouse gain across three target sizes. ECR: m. extensor carpi radialis, ED: m. extensor digitorum superficialis, ECU: m. extensor carpi ulnaris, FCR: m. flexor carpi radialis, RTRAP and LTRAP: right and left m. trapezius, NE: right neck extensor muscle (upper part of m. trapezius and m. splenius capitis). (A) The young group, (B) the elderly group. EMG was reported from the session where working speed and thereby working load was kept constant by guidance of a metronome (50 clicks/min).
Finally, the values for NE were 4.1% and 9.9%EMGmax for the young and elderly, respectively.
4. Discussion The present study demonstrated a rather complex relationship between computer MG and performance. Thus, the young subjects seem to have an optimum at MG 1:4 provided that the target sizes were large or medium. This optimum was not found among the elderly. Additionally, at the highest MG a decrease in working speed and hit rate was found for both age groups, especially when the target size was small. Thus, our hypothesis regarding the relation between performance and motor demand was confirmed for the small targets but the interaction between the two parameters turned out to be more complex than originally suggested
ARTICLE IN PRESS J. Sandfeld, B.R. Jensen / Applied Ergonomics 36 (2005) 547–555
and hence, the hypothesis was not confirmed for the two largest targets. Still, the important notion can be made, that reducing MG may enhance performance when working with small screen objects. Because the MG determines the distance the computer mouse must be moved between targets, it may be expected that MG is related to the EMG-level. In both groups, the ED muscle activity and in the young group, the ECR and FCR muscle activity increased with increasing MG. However, the absolute differences in muscle activity across the three MG were small and probably without major functional relevance. This indicates that the activation of the forearm muscles may be determined to a higher degree by the demand for stability of the forearm/hand rather than by the movement of the hand per se. Furthermore, the combination of a high MG and small targets had a higher impact on performance in the elderly group compared to the young group. This may be explained by several factors. Working with the computer mouse utilizes mainly closed loop movement cycles, where movement causes afferent feed-back information which is processed in the brain and integrated with new motor programs, resulting in new movements. Possible factors involved in the reduced hand-eye control generally found in elderly are several. Morphological age-related changes such as death of neurons innervating the large type II muscle fibers, cause a reorganization of the motor unit composition resulting in enlargement of type I motor units (Doherty et al., 1993; Lexell, 1995). Larger type I motor units make it more difficult to graduate the force output at low intensity work like mouse handling. Age-related changes in neurological factors such as decline in the number of neurons (Vandervoort, 2002), reduction in motor unit excitability (Sabbahi and Sedgwich, 1982) and slowing of nerve conduction velocity (Dorfman and Bosley, 1979) also contribute to the decrease in motor control in elderly even though the impact of each of these factors is very difficult to quantify. Finally, slower information processing and motor programming in the brain which has been found in the elderly may be an influential factor, since a longer duration of each closed loop cycle reduces the number of corrections possible within a fixed time frame. Consequently, the execution of a task will either be less precise or take longer time for an elderly compared to a young person (Cerella, 1985; Rubichi et al., 1999; Van der Lubbe and Verleger, 2002). The generally higher EMG-level in the elderly compared to the young found in the present study has also been found in earlier studies using the multidirectional pointing task and it is speculated whether this phenomenon may be the result of a higher central drive and/or a higher antagonist activity in the elderly. It has been shown that overall higher EMG-level increase joint stiffness which stabilize the movement and reduce
553
the acceleration of movements (Milner and Cloutier, 1998; Seidler-Dobrin et al., 1998). Thus, in contrast to the above mentioned factors the higher EMG-level in the elderly may be a compensatory mechanism which in part counteracts the decline in motor control. As the reported RMS-EMG is the percentage of the maximal EMG, it could be speculated if the higher relative EMGlevel in the elderly is a result of a reduced ability of the elderly to activate the muscle maximally during the MVC. However, several studies addressing this topic show that the active elderly are able to fully activate the available muscle mass (Connelly et al., 1999; De Serres and Enoka, 1998; Jakobi and Rice, 2002; Klein et al., 2001; Phillips et al., 1992). In spite of between group differences in muscle activity level, similar activation pattern between muscles were found in the two age groups. Finally, we found that concentrated mouse work demands a relatively high level of muscle activity in especially ED and ECU, but also in the NE in the elderly group. When considering the often long duration of every day mouse work, this may be a factor contributing to development of work related muscular skeletal symptoms. According to our second hypothesis we did not expect to find a relationship between visual demand and performance. However, the present results indicate that increased visual demand reduced performance even though index of difficulty was kept constant. This may be explained by the limitations of the human visual acuity, which among other things depends on the area of the retina stimulated. With decreasing target size the area of the retina stimulated was decreased and hence, a reduction of the ratio between the visual signal and the uncertainty in the visual system occurred. That is, there was a reduction in the signal-to-noise ratio which caused a decline in quality of the visual object projected in the visual cortex. However, the average person with normal vision can resolve points separated by one minute of arc (¼ 1=60 of one degree) in a traditional Snellen test but the size of the smallest screen target (2.5 2.5 mm) in the present study equalled approximately 14 min of arc and should thus, in theory not constitute any visual difficulties (Davson, 1990). However, the task was to place the cursor within the target and thus the subjects had to visually detect space between the cursor tip and the edge of the target, which substantially increases the visual demand. Furthermore, when working at the computer screen and performing a multi-directional pointing task, visual acuity is influenced by additional factors which potentially increases the demand on the visual system compared to the Snellen test. For instance, in contrast to the Snellen test where the subject is situated 6 m from the chart, the image at the screen is located at a near or intermediate distance of approximately 60 cm which means that accommodation and vergence may play a role in maintaining clarity of the
ARTICLE IN PRESS 554
J. Sandfeld, B.R. Jensen / Applied Ergonomics 36 (2005) 547–555
image (Kurimoto et al., 1983). The time factor also influences the performance because time is needed for binocular refixation whenever a new target appeared (Adrian, 2003; Kaufman et al., 1992). Finally, factors such as the degree of illumination and contrast which has a great influence on visual acuity may have been different in the present task compared to the settings needed for testing with the Snellen chart. Taken together, the computer task may have constituted a higher visual demand than what could be expected based merely on the screen target sizes. Earlier studies of motor control related to computer work have found a reduction in performance in the elderly compared to the young and specifically, it has been shown that even though elderly have the same hit rate during a multi-directional pointing task compared to young, this outcome is attained on the expense of the working speed which is lower in the elderly. However, when a demand for a certain clicking frequency was applied to the clicking task, a reduced hit rate has been found in the elderly compared to the young. Also, it has been found that increasing precision demand reduces the performance of the elderly to a larger extent than the young (Laursen et al., 2001; Smith et al., 1999). The present findings confirmed these earlier results as the working speed was lower in the elderly group compared to the young, while no difference in hit rate was found. Fitts’ index of difficulty have been widely used in the human-computer interaction literature and recommended for comparison of usability of non-keyboard input devices. The index of difficulty is based on the assumption that target width and inter-target distances are related in a manner based on information technology (Fitts, 1954). Fitts’ index of difficulty was kept constant in all trials in the present study and therefore, one should expect a similar level of performance in all the trials. However, the data showed major variations in the performance, depending on the combination of mouse gain and target size. One likely explanation for these findings is that Fitts’ index of difficulty was originally validated in a one-to-one setting where the amplitude of the arm/hand movement was directly related to the visual input, which is rare during use of computer input devices. Based on this, it is recommended that usability evaluations and comparisons of input devices using the ISO-standard, which is an evolution of the Fitts equation should include the gain of the pointing device as an additional parameter that alters task difficulty. Otherwise wrong conclusions regarding usability of the pointing device may be drawn.
5. Conclusion The study contributes to the understanding of the interaction between the physical demands involved in
use of computer mouse, performance and muscle activation in relation to aging. In both age groups performance was reduced with increasing visual demands, indicating that computer mouse work with small screen objects and letters should be avoided if possible. The combination of small target sizes and high mouse gain reduced performance severely and this was especially pronounced in the elderly group. Thus, if working with small screen objects or letters, it is recommended to decrease mouse gain. A mouse gain of 1:4 seems to be optimal for young computer users. The muscle activation levels were in general higher for the elderly than the young group, although the activation strategy was similar. Finally, it is recommended to expand the guidelines for testing input device usability, to include pointing device gain as task difficulty is dependent on input device gain.
Acknowledgements We would like to thank the Danish Research Agency for supporting the project. Also we would like to acknowledge the technical assistance of Marianne Pilegaard, Institute of Exercise and Sport Sciences and Bjarne Laursen, National Institute of Public Health. References Adrian, W., 2003. The effect of observation time and contrast on visual acuity. Clin. Exp. Optometry 86 (3), 179–182. Cerella, J., 1985. Information-processing rates in the elderly. Psychol. Bull. 98 (1), 67–83. Connelly, D.M., Rice, C.L., Roos, M.R., Vandervoort, A.A., 1999. Motor unit firing rates and contractile properties in tibialis anterior of young and old men. J. Appl. Physiol. 87 (2), 843–852. Cooke, J.D., Brown, S.H., Cunningham, D.A., 1989. Kinematics of arm movements in elderly humans. Neurobiol. Aging 10, 159–165. Davson, H., 1990. Physiology of the Eye. Macmillan Press, London. De Serres, S.J., Enoka, R.M., 1998. Older adults can maximally activate the biceps brachii muscle by voluntary command. Am. Physiol. Soc. 84 (1), 284–291. Doherty, T.J., Vandervoort, A.A., Taylor, A.W., Brown, W.F., 1993. Effects of motor unit losses on strength in older men and women. J. Appl. Physiol. 74 (2), 868–874. Dorfman, L.J., Bosley, T.M., 1979. Age-related changes in peripheral and central nerve conduction in man. Neurology 29, 38–44. Fitts, P.M., 1954. The information capacity of the human motor system in controlling the amplitude of movement. J. Exp. Psychol. 47 (6), 381–391. International Organization for Standardization, 1997. ISO 9241, Part 9 CD: Non-keyboard Input Devices, 1–47. Jakobi, J.M., Rice, C.L., 2002. Voluntary muscle activation varies with age and muscle group. J. Appl. Physiol. 93 (2), 457–462. Kaufman, P.L., Adler, F.H., Hart, W.M., Moses, R.A., 1992. Adler’s physiology of the eye. Mosby Year Book, St. Louis. Klein, C.S., Rice, C.L., Marsh, G.D., 2001. Normalized force, activation, and coactivation in the arm muscles of young and old men. J. Appl. Physiol. 91 (3), 1341–1349.
ARTICLE IN PRESS J. Sandfeld, B.R. Jensen / Applied Ergonomics 36 (2005) 547–555 Kuorinka, I., Jonsson, B., Kilbom, A˚., Vinterberg, H., BieringSørensen, F., Andersson, G., Jørgensen, K., 1987. Standardised Nordic questionnaires for the analysis of musculoskeletal symptoms. Appl. Ergonomics 18 (3), 233–237. Kurimoto, S., Iwasaki, T., Nomura, T., Noro, K., Yamamoto, S., 1983. Influence of VDT (visual display terminals) work on eye accommodation. J.UOEH 5 (1), 101–110. Laursen, B., Jensen, B.R., Ratkevicius, A., 2001. Performance and muscle activity during computer mouse tasks in young and elderly adults. Eur. J. Appl. Physiol. Occup. Physiol. 84 (4), 329–336. Laursen, B., Jensen, B.R., Garde, A.H., Jørgensen, A.H., 2002. Effect of mental and physical demands on muscular activity during the use of a computer mouse and a keyboard. Scand. J. Work Environ. Health 28 (4), 215–221. Lexell, J., 1995. Human aging, muscle mass, and fiber type composition. J. Gerontol. 50A (Special Issue), 11–16. Milner, T.E., Cloutier, C., 1998. Damping of the wrist joint during voluntary movement. Exp. Brain Res. 122 (3), 309–317.
555
Phillips, S.K., Bruce, S.A., Newton, D., Woledge, R.C., 1992. The weakness of old age is not due to failure of muscle activation. J. Gerontol. 47 (2), M45–M49. Ranganathan, V.K., Siemionow, V., Sahgal, V., Yue, G.H., 2001. Effects of aging on hand function. J. Am. Geriatr. Soc. 49 (11), 1478–1484. Rubichi, S., Neri, M., Nicoletti, R., 1999. Age-related slowing of control processes: evidence from a response coordination task. Cortex 35 (4), 573–582. Sabbahi, M.A., Sedgwich, E.M., 1982. Age-related changes in monosynaptic reflex excitability. J. Gerontol. 37 (1), 24–32. Seidler-Dobrin, R.D., He, J., Stelmach, G.E., 1998. Coactivation to reduce variability in the elderly. Motor Control 2, 314–330. Smith, M.W., Sharit, J., Czaja, S.J., 1999. Aging, motor control, and the performance of computer mouse tasks. Human Factors 41 (3), 389–396. Van der Lubbe, R.H.J., Verleger, R., 2002. Aging and the Simon task. Psychophysiology 39 (1), 100–110. Vandervoort, A.A., 2002. Aging of the human neuromuscular system. Muscle Nerve 25 (1), 17–25.