Effects of computer mouse control-display gain on upper extremity muscle fatigue,subjective fatigue and user performance

Abstract

Background

Operating a mouse for long periods of time increases the risk of muscle fatigue in the upper limbs of office workers.

Objective

This study aimed to contribute to the field of knowledge by providing guidance on the selection of a suitable computer mouse control-display gain (CDG) for office workers, and to reduce muscle fatigue and fatigue accumulation in office workers who use computer mice.

Method

Thirty healthy office workers used five different control-display gain settings to complete five sessions of a 30-min Fitts’ task. Muscle activity in the extensor digitorum, extensor carpi ulnaris and flexor digitorum superficialis were measured by surface electromyography (sEMG). Subjective muscle fatigue was assessed using the Borg Rating Scale and user performance was recorded using the GoFitts program.

Results

It was found that when the control-display gain was set to 12.97, the highest effective frequency of operation was reached by the users. Additionally, the subjects reported more subjective muscle fatigue at a control-display gain of 12.97, and the sEMG data showed that manipulation led to more upper limb muscle fatigue when the control-display gain was set to 8.64 and 12.97. Upper limb muscle fatigue, subjective muscle fatigue and user performance also varied between genders. Overall, female subjects had lower error rates while male subjects reported higher levels of subjective muscle fatigue and were more adaptable to the low control-display gain setting.

Conclusion

This study found a correlation between muscle fatigue and subjective fatigue with control-display gain. Similar to the subjective fatigue for control-display gain settings, the control-display gain settings that are commonly used by office workers may lead to higher levels of fatigue in the forearm muscles. This study could provide guidance for the selection of computer mouse control-display gain to reduce muscle fatigue and fatigue accumulation in office workers who use computer mice.

Keywords

muscle fatigue computer mouse control-display gain user performance ergonomics electromyography

Introduction

In current society, computers have become the most commonly used piece of electronic work equipment. As computer hardware technology continues to develop, computer input and output devices are also constantly updated. Since the birth of the first commercial mouse in 1981, the function and modeling of computer mice has matured over time. However, there have not been many specific studies on the effects of mouse control-display gain (CDG) settings on user performance and muscle fatigue. The typical approach for office workers is to adapt to the default settings of hardware devices, rather than proactively select the most optimal and efficient settings.

In recent years, there has been considerable research into work-induced musculoskeletal discomfort.^1–4 Research on the ergonomics of computer input devices, such as the mouse,⁵ controller,^6,7 touchscreen,⁸ sidestick,⁹ and touch pen¹⁰ has gained more attention as the normalized use of computers has been associated with an increase in physical injuries resulting from prolonged computer work.^11,12 Intensive mouse use has been associated with an increased risk of musculoskeletal disorders of the upper extremities, including carpal tunnel syndrome.^13–15 A study in the 1990s revealed increased carpal tunnel pressure during mouse use, and prolonged, intensive mouse work may increase the risk of median neuropathy.¹⁶ The risk of median neuropathy has been acknowledged in the past, but those previous studies lack strong evidence of a causal relationship between pointing devices and the development of carpal tunnel syndrome (CTS).¹⁷ However, recent studies have confirmed the potentially high prevalence of musculoskeletal discomfort or fatigue resulting from the use of traditional visual display terminals (VDTs).^18,19 It is encouraging to note that a number of government departments have devised guidelines to assist in the configuration of workstations, with an aim of minimizing the risk of musculoskeletal discomfort.²⁰ Despite this progress, the current body of research still has not revealed an association between CDG settings and muscle fatigue or subjective physical discomfort.

CDG is a unitless coefficient that maps the physical movement of the pointing device to the movement of the display pointer on screen. In this study, CDG is expressed as the ratio of the distance the pointer moves on the screen to the distance the mouse moves on the mouse pad. The CDG value can be calculated by dividing the velocity of the pointer by the velocity of the device. Research on CDG started in the 1990s and has been gaining attention in recent years. Topics of research changed from on variable-gain mice initially, to the current focus on the effects of different CDGs. Although initial studies showed that changing CDG had little effect on performance,²¹ subsequent studies have confirmed that different populations prefer different CDG settings and have different levels of performance at different CDGs.^22,23 A study conducted more than a decade ago²⁴ showed that users have different preferences for CDG depending on the screen size, and a formula was provided for calculating the available CDG range. Such studies have analyzed user preferences or performance at different computer mice CDG settings, but they have not examined the relationship between CDG and possible muscle fatigue. In recent years, researchers have focused on how muscle fatigue is affected by CDG with new types of controllers, especially during virtual reality interactions. One study on virtual reality suggests that to avoid neck muscle fatigue, the user should avoid placing the target at 15° above and 30° below eye level.²⁵ Another study on virtual reality explored the recommended CDG values when using different body parts. It concluded that a CDG value of 1.7 should be avoided in shoulder-based interactions to improve user performance.²⁶ Furthermore, a study on CDG settings in gesture control ascertained the importance of maximizing free-hand performance.²⁷ Recently, researchers have thoroughly examined the effects of CDG settings in virtual reality interactions on user performance and muscle fatigue. However, research on CDG in conventional mice cannot be ignored, and studies on the effects of CDG on muscle fatigue are still relatively scarce.

As early as 2004, a study was conducted to evaluate the design of computerized mice using Surface electromyography (sEMG) to collect data during static grasping.²⁸ This study found that a well-designed mouse was effective in reducing the activity level of the upper limb muscles, taking into consideration functional parameters, wrist posture deviations and comfort level. A comparison between a computer mouse and a touch pen was conducted using sEMG to study fatigue caused by distal pointing²⁹ where the median frequency (MDF) slope was used to represent the muscle fatigue index. The study found that pointing methods operated by different body parts led to different levels of fatigue. and the level of fatigue was not related to the CDG of the distal pointing devices. Subsequent researchers have used sEMG to conduct more detailed studies of muscle activity during mouse use. These studies have found that the shape and weight of the mouse, as well as the CDG settings, have an effect on both muscle activity and user performance.^30–32 Additionally, a recent study has indicated that mouse clicking may not result in performance fatigue. However, high levels of extensor activity may explain common injuries among gamers.³³ Such studies tend to analyze the effects of the mouse or other input devices on muscle activity in order to guide future mouse design. However, the studies mentioned above did not address the possible reduction in fatigue that may be achieved by adjusting the existing CDG mouse settings.

In summary, the previous studies did not explore the potential for improving muscle fatigue by changing mouse CDG settings, nor have they examined user performance with different CDG settings. The aim of this study is to contribute to the field of knowledge by providing guidance on the selection of computer mouse CDG to reduce muscle fatigue and fatigue accumulation in office workers. The possibility of reducing office workers’ muscle fatigue by adjusting the CDG value was explored by comparing muscle fatigue and user performance under different CDG settings. This investigation would add to the current knowledge on the effects of CDG settings on upper limb muscle fatigue and user performance. The specific objectives of this study were to induce fatigue in the upper limb muscles of the subjects performing Fitts’ Tasks at different CDG values, such that upper limb muscle fatigue data at different CDG values could be obtained. The results of this study may provide a reference for office workers to optimize their CDG values, which has practical implications for reducing upper extremity fatigue and occult diseases caused by office work.

Method

Participants

A convenience sample of college-aged participants was recruited to participate in this exploratory study. G*power software was used to analyze the sample size and it was found that the sample size was at least 22 below the recommended sample size for the experimental setup of this study. Therefore, the number of subjects was increased to improve data robustness. The subjects were screened by carpal tunnel syndrome 6 to ensure the exclusion of patients with carpal tunnel syndrome and other subjects with musculoskeletal complaints. A total of 30 right-hand-dominant participants (15 males and 15 females) were recruited after the screening process was completed. All participants were free from any type of musculoskeletal disorders and had no occurrence of notable pain or injury in the neck, shoulder, back, buttocks, or extremities for at least 6 months. Additionally, the recruits reported using the mouse for more than 10 h of work per week in the six months prior to the experiment. The participants’ ages ranged from 22 to 31 years (males: 24.7 ± 1.7; females: 24.1 ± 1.5; years). The effect of age on the study population was explored in preparation for future expansion of subjects. All subjects in the study were asked to complete a questionnaire prior to participation. Demographic data, including age and gender, are shown in Table 1. The experimental procedure was explained to each subject and informed consent was obtained before the experiment began. This study is in compliance with the tenets of the Declaration of Helsinki and was approved by the Ethics Committee of the Northwestern Polytechnical University (IRB No. are 201502024 and 202002024).

Table 1.

Demographic data of subjects.

	Female (n = 15)	Male (n = 15)	Total
Age (year) [mean (SD; range)]	Mean = 24.07	Mean = 24.67	Mean = 24.37
Age (year) [mean (SD; range)]	(1.49; 22–26)	(1.72; 23–28)	(1.61; 22–28)
Body height (cm) [mean (SD; range)]	Mean = 164.20	Mean = 174.67	Mean = 169.43
Body height (cm) [mean (SD; range)]	(4.18; 153–172)	(4.40; 168–182)	(6.79; 153–182)
Body weight (kg) [mean (SD; range)]	Mean = 52.20	Mean = 73.20	Mean = 62.70
Body weight (kg) [mean (SD; range)]	(4.86; 47–65)	(12.64; 55–109)	(14.23; 47–109)
Body Mass Index [mean (SD; range)]	Mean = 19.36	Mean = 23.90	Mean = 19.36
Body Mass Index [mean (SD; range)]	(1.54; 17.57–22.49)	(3.28; 19.03–33.27)	(1.54; 17.57–22.49)

Apparatus

A high-performance laptop (CPU: Intel i7-12700H) with a GeForce RTX 3070 Laptop GPU (graphics processing unit) and a 1920 × 1080 resolution LCD (Liquid crystal display) monitor was used for this study (Windows 11 operation system). A Logitech G304 (see Figure 1(a)) wireless mouse was chosen as the experimental mouse. The devices listed above were selected to minimize system lag and physical interference, and to emulate the display and input devices that are commonly used by office workers.

Figure 1.

Experimental mouse and tested muscles: (a) The experimental mouse (Logitech G304) (b) Electrode placement on three forearm muscles: extensor digitorum (ED), flexor digitorum superficialis (FDS), and extensor carpi ulnaris (ECU).

Five different CDG settings (2.16, 4.23, 8.64, 12.97, 21.62) were tested and the testing order was randomized for each participant. The settings were determined based on the CDG values of the experimental mouse at five different Dots Per Inch (DPI) settings (200, 400, 800, 1200, and 2000). The DPI was adjusted using the Logitech G HUB driver software for the mouse. Pixels per inch (PPI) is a measurement used to describe the resolution of a display screen while DPI is a standard used to measure mouse sensitivity. DPI indicates the number of movements detected by the sensor for every inch of movement of the device. Under the default settings of Windows, the pointer moves one pixel on the screen for every movement detected by the mouse. The formula for calculating CDG using screen PPI and mouse DPI is as follows²³:

C D G = \frac{D P I}{P P I}

(1)

User performance was tested using the “GoFitts” software³⁴(see Figure 2) and quantified by Error Rate (ER), Movement Times (MT) and Effective Operation Frequency (EOF). ER is defined as the ratio of the number of successful operations performed by the subject to the total number of operations performed. MT is directly recorded by the Gofitts program and refers to the duration between clicking two successive targets, and the average MT was calculated based on all clicks for each protocol. EOF is defined according to the Throughput form in the Fitts’ Task,³⁵ and describes and assesses the efficiency with which subjects perform effective input in an experimental task. The EOF is calculated by:

E O F = \frac{1 - E R}{M T}

(2)

Figure 2.

Gofitts’ program analysis interface.

Subjective muscle fatigue³⁶ was assessed using the Borg Rating of Perceived Exertion Category Ratio CR-10 Scale.³⁷ In this study, subjective muscle fatigue was defined as the degree of fatigue, felt by the subjects in their upper limb muscles after performing the experimental tasks. Participants rated their level of muscle fatigue on a scale of 1 to 10, with 0 indicating no exertion at all and 10 indicating maximal exertion that prevented the completion of the experiment.

The operation of a computer mouse for selective tasks can be subdivided into two main types of actions, which are clicking the buttons and moving the mouse. These actions are performed by different muscle groups, including the extensor digitorum, flexor digitorum superficialis, extensor pollicis brevis, flexor carpi radialis, extensor carpi radialis, extensor carpi ulnaris, and the first dorsal flexor carpi radialis interosseus. Based on recent findings in similar studies, extensor digitorum (ED), extensor carpi ulnaris (ECU), and flexor digitorum superficialis (FDS) are the main representative muscles.^22,33,38,39 Upper limb muscle fatigue was measured using sEMG and processed as mean power frequency (MPF) for analysis. sEMG is commonly used to evaluate the muscle activity required to perform repetitive hand motions.^28,40 The sEMG signals were captured using an electromyogram amplifier (model: EMG100C) manufactured by BIOPAC. Three groups of electrode pads were placed on the subject's ED, ECU and FDS to capture sEMG signals (see Figure 1(b)). The sampling frequency was set to 1000 Hz then filtered at a bandwidth of 10 to 450 Hz.⁴¹

Experimental procedure

Before collecting the data, the experimental procedure and requirements of the Fitts’ tasks were explained in detail to all participants. The participants adjusted the chair and table height to their preferred level and positioned the keyboard, computer mouse, and monitor according to their preferences. The room temperature was maintained at 25°C. Each experiment consisted of five sessions, and each session contained five tasks with a duration of 5 min each. A total of 750 tasks were performed during the study (see Figure 3). After providing informed consent, the participants completed a baseline survey to collect demographic information and gaming experience.

Figure 3.

Flowchart of experimental steps.

The Fitts’ tasks measured performance at each CDG setting. After the participant adjusted the height of the table, chair, and screen, the experimenter instructed the participant to complete a Fitts’ task by clicking and moving the mouse. The experimenter then asked the participant to continue performing Fitt's tasks for several minutes until the participant became familiar with the CDG settings of the experiment. This process took between 2 to 10 min, depending on the CDG settings and individual participants. The participants were instructed to move the mouse to position the cursor over a highlighted circular target. They then had to click and release the left button and move the cursor to the next highlighted target. The participants were asked to perform this task as rapidly and accurately as possible. A sequence consisted of clicking on 4 circles with 2 ‘amplitudes’ (center to center distance of 400,800 pixels), and 2 ‘widths’ (target radius size of 20, 40 pixels). A total of 256 movements and clicks were executed in each sequence. The program randomly generated every possible combination of amplitudes and widths (GoFitts’). A click outside the highlighted circular target was defined as an error and a beep was generated to alert the participant of the error. The participant proceeded to the next target without correcting the error. The mouse resolutions were set to 200, 400, 800, 1200, 2000 DPI during the performance of the five different sessions of Fitts tasks. The order of the five sessions was randomized and counterbalanced to avoid order effects. The mouse DPI indicates the number of dots or pixels that are sent by the mouse to the computer per inch of travel. The higher the mouse DPI value, the higher the sensitivity of the mouse, and the farther the cursor will move when the mouse moves the same distance. In addition, the mouse pointer speed was kept the same for all participants. Mouse pointer speed is identical to DPI in that a higher pointer speed means the cursor will travel more pixels for every inch the mouse moves.

Each participant performed five sessions of testing. For each session, five sets of 5-min Fitts’ tasks were performed (see Figure 4). Between each set of Fitts’ tasks, the participants rated their muscle fatigue using a Borg rating scale and rested for one minute.

Figure 4

Distribution of user performance metrics at different CDG values: (a) Interval plot showing differences in MT distributions at different CDG settings and 95% confidence intervals for the means. (b) Interval plot showing differences in ER distributions at different CDG settings and 95% confidence intervals for the means. (c) Interval plot showing differences in EOF distributions at different CDG settings and 95% confidence intervals for the means.

Data analysis

The sEMG data were preprocessed and converted to MPF using MATLAB R2018b. Firstly, the raw EMG data were subjected to full-wave rectification and band-pass filtering from 10–450 Hz. Then the 25-min experimental data from each group of experiments were divided into 25 segments on average, and the MPF values of the windows were calculated separately with a time window of 60 s. Next, the MPF values were linearly regressed to find the slope of the MPF. The MPF was calculated as follows⁴²:

M P F = \frac{\int_{f_{1}}^{f_{2}} f P (f) d f}{\int_{f_{1}}^{f_{2}} P (f) d f}

(3)

In the above equation, f denotes the frequency of the signal and P(f) denotes the power spectral density function of the signal. Statistical work was performed using Orgin2019 to compare the MT, ER, EOF, MPF and fatigue scores. Paired Wilcoxon Signed-Rank tests were used to compare subjective fatigue, user performance, and muscle activity between the different CDG settings, with P < .05 considered statistically significant. The Body Mass Index (BMI) of the participants was tested for normality using the Shapiro-Wilk test, and Spearman's correlation analysis and partial correlation analyses were performed on BMI with MT, ER, EOF, MPF slope and Borg rating scale data.

Results

A total of 30 individuals participated in the study (Table 1), and data was gathered on user performance (Table 2), sEMG upper extremity muscle activity (Table 3), and subjective muscle fatigue (Table 4). The results showed that different CDG values led to differences in user performance, upper limb muscle activity and subjective muscle fatigue. Furthermore, user performance varied significantly between genders at different CDG settings.

Table 2.

Mean values of MT, ER and EOF at different CDG values and their standard deviations for subjects of different genders.

CDG	Muscles	Female		Male		n	p
CDG	Muscles	Mean	SD	Mean	SD	n	p
2.16	MT	1042.62	151.31	1026.94	182.43	0.29	0.78
	ER	0.03	0.02	0.04	0.02	−1.78	0.10
	EOF	0.95	0.14	0.96	0.16	−0.22	0.83
4.32	MT	967.71	124.45	877.59	129.95	1.63	0.12
	ER	0.03	0.03	0.05	0.03	−1.98	0.07
	EOF	1.02	0.14	1.10	0.15	−1.35	0.20
8.64	MT	912.18	136.29	814.54	128.06	1.79	0.10
	ER	0.04	0.05	0.06	0.04	−1.53	0.15
	EOF	1.08	0.17	1.18	0.19	−1.44	0.17
12.97	MT	874.09	99.31	759.53	98.64	3.06	0.01*
	ER	0.04	0.04	0.09	0.05	−2.90	0.01*
	EOF	1.12	0.12	1.23	0.18	−2.00	0.07
21.62	MT	887.69	111.30	765.02	80.89	3.17	0.01*
	ER	0.04	0.05	0.09	0.05	−2.45	0.03*
	EOF	1.09	0.13	1.20	0.16	−2.02	0.06

Note. CDG, Control-display gain; MT, movement times; ER, error rate; EOF, effective operation frequency.

Table 3.

Mean MPF slopes and their standard deviations for different muscles in subjects of different genders at different CDG values.

CDG	Muscles	Female		Male		Z	W
CDG	Muscles	Mean	SD	Mean	SD	Z	W
2.16	ED	−0.04	0.26	0.01	0.24	−1.16	0.25
	ECU	−0.03	0.29	0.14	0.38	−1.51	0.14
	FDS	−0.03	0.29	0.14	0.38	−1.51	0.14
4.32	ED	0.10	0.28	−0.07	0.20	1.79	0.07
	ECU	0.03	0.35	0.05	0.26	−0.60	0.56
	FDS	0.15	0.44	0.11	0.54	0.26	0.80
8.64	ED	−0.08	0.22	−0.12	0.22	0.26	0.80
	ECU	0.03	0.30	−0.01	0.27	0.26	0.80
	FDS	−0.01	0.30	0.18	0.42	−1.22	0.23
12.97	ED	−0.05	0.19	−0.10	0.21	0.31	0.76
	ECU	0.04	0.40	0.03	0.23	0.00	1.00
	FDS	0.08	0.89	0.17	0.86	−0.14	0.89
21.62	ED	0.05	0.27	−0.06	0.15	1.73	0.08
	ECU	−0.03	0.37	0.06	0.27	−1.05	0.30
	FDS	0.28	0.76	0.25	0.63	−0.37	0.72

Note. MPF, mean power frequency; CDG, Control-display gain; ED, extensor digitorum; ECU, extensor carpi ulnaris; FDS, flexor digitorum superficialis.

Table 4.

Mean Borg Rating and their standard deviations for different muscles in subjects of different genders at different CDG values.

CDG	Muscles	Female		Male		n	p
CDG	Muscles	Mean	SD	Mean	SD	n	p
2.16	ED	2.93	1.75	3.63	1.73	−1.03	0.32
	ECU	1.53	1.20	2.60	1.51	−1.84	0.09
	FDS	1.80	1.58	1.90	1.40	−0.16	0.88
4.32	ED	2.80	1.65	2.83	1.62	−0.10	0.92
	ECU	1.67	1.37	2.33	1.68	−1.28	0.22
	FDS	1.67	1.05	1.90	1.44	−0.53	0.60
8.64	ED	3.30	1.24	2.97	1.88	0.71	0.49
	ECU	2.23	1.45	2.47	1.92	−0.41	0.69
	FDS	1.53	1.23	2.13	1.60	−1.40	0.18
12.97	ED	2.97	1.68	2.50	1.44	0.77	0.45
	ECU	1.67	1.03	1.73	1.29	−0.17	0.87
	FDS	1.77	1.05	1.70	1.13	0.20	0.84
21.62	ED	3.33	1.99	2.73	1.58	1.09	0.29
	ECU	1.80	1.39	1.97	1.32	−0.31	0.76
	FDS	1.87	1.23	1.87	1.09	0.00	1.00

Note. CDG, Control-display gain; ED, extensor digitorum; ECU, extensor carpi ulnaris; FDS, flexor digitorum superficialis.

User performance

User performance varied significantly for different CDG values, as evidenced by differences in ER, MT and EOF. Specifically, increases in CDG values led ER to increase (see Figure 4(b)), while MT decreased, reaching a minimum at a CDG value of 12.97 (see Figure 4(a)).

Conversely, EOF reached a maximum when CDG was set to 12.97 (see Figure 4(c)). The EOF value was lower at a CDG value of 21.62 due to an increase in the ER despite the shorter MT.

Upper extremity muscle fatigue

The distribution of MPF slope in ED shows that a smaller slope was reached at CDG values of 8.64 and 12.97. For ECU, there was no significant correlation between the MPF slope and CDG. However, in FDS, the MPF slope increased as CDG value increased. Overall, the MPF slope of ED was smaller than that of ECU, while the MPF slope of ECU was smaller than that of FDS (see Figure 5).

Figure 5.

Interval plot showing the difference in the distribution of s for different muscles at different CDG settings, with 95% confidence intervals for the mean values.

Borg Fatigue Scale

From the statistics of the Borg Fatigue Scale, ED had higher fatigue scores than ECU and FDS (P < .05). The ECU fatigue score was higher at a CDG value of 8.64 than with a CDG value of 12.96 (P < .05). At CDG values other than 8.64, the ECU fatigue score was similar to that for FDS. There were no significant relationships between FDS fatigue score and CDG values. All mean fatigue scores ranged from 1 to 4 (see Figure 6).

Figure 6.

Interval plot showing the difference in the Borg Fatigue score distribution for different muscles at different CDG settings, with 95% confidence intervals for the mean values.

Gender difference

Gender differences did not affect the general trend of user performance across different CDG values. However, for all CDG values greater than 2.16, the MT of males was generally smaller than the MT of females and the ER of females was smaller than the ER of males. Both of these gaps widened as CDG values increased. The ER distribution shows that the error rate for females was not significantly related to the CDG value since the distribution remained similar regardless of changes in CDG settings. Finally, the distribution plot for gender difference in EOF shows that the EOF of males was higher than that of females, except for when CDG was at 2.16. The EOF for both genders was highest at a CDG value of 12.97 (see Figure 7).

Figure 7.

Distribution of gender difference in the distribution of MPF slopes for different muscles at different CDG settings, with 95% confidence intervals for the mean values: (a) Interval plot showing gender differences in MT distributions at different CDG settings and 95% confidence intervals for the mean values. (b) Interval plot showing gender differences in ER distributions at different CDG settings and 95% confidence intervals for the means. (c) Interval plot showing gender differences in EOF distributions at different CDG settings and 95% confidence intervals for the mean values.

Overall, there were no obvious patterns in the distribution of MPF slopes across genders. However, it was found that at CDG values of 4.32 and 21.97, the MPF slopes of female EDs were higher than the MPF slopes of male EDs. At a CDG value of 2.16, the MPF slope for female ECUs was lower than the MPF slope for male ECUs. The MPF slopes of female FDS were lower than the MPF slopes of male FDS at CDG values of 2.16 and 8.64. In summary, the FDS and ECU of females were more susceptible to fatigue than the FDS of males for the same number of working hours, while the ED of males was more susceptible to fatigue than that of females (See Figure 8).

Figure 8.

Distribution of gender differences in user performance metrics for different CDG values: (a) Interval plot showing the difference in the distribution of MPF slopes for ED at different CDG settings, with 95% confidence intervals for the mean values. (b) Interval plot showing the difference in the distribution of MPF slopes for ECU at different CDG settings, with 95% confidence intervals for the mean values. (c) Interval plot showing the difference in the distribution of MPF slopes for FDS at different CDG settings, with 95% confidence intervals for the mean values.

The difference in subjective muscle fatigue in ED across genders was more interesting. When the CDG values were low, subjective muscle fatigue in female ED was lower than subjective muscle fatigue in male ED. When CDG values were high, the results were reversed, with subjective muscle fatigue in female ED being higher than subjective muscle fatigue in male ED. When the CDG value was at 4.32, there were no significant differences in the ED fatigue scores between males and females. When the CDG value was less than 8.64, subjective muscle fatigue scores were lower in female ECUs than in males. In all other cases, there were no significant differences in the subjective muscle fatigue distribution of ECU between males and females. FDS fatigue scores were higher in males when the CDG value was 8.64, but there were no significant differences in the FDS fatigue scores between males and females at other CDG values (see Figure 9).

Figure 9.

Distribution of gender differences in Borg Fatigue scores for different CDG values: (a) Interval plot showing the difference in the distribution of Borg Fatigue scores for ED at different CDG settings, with 95% confidence intervals for the mean values. (b) Interval plot showing the difference in the distribution of Borg Fatigue scores for ECU at different CDG settings, with 95% confidence intervals for the mean values. (c) Interval plot showing the difference in the distribution of Borg Fatigue scores for FDS at different CDG settings, with 95% confidence intervals for the mean values.

Correlation analysis between BMI and other experimental results

The Shapiro-Wilk test was performed on the BMI data of the subjects, and the BMI distribution of the subjects was found to be non-normal. The Spearman correlation analysis was then conducted and it was found that BMI was weakly correlated with MT and ER when CDG values ranged between 4.32 to 21.62, and BMI was weakly correlated with EOF at CDG values of 4.32, 12.97 and 21.62 (see Table 5). Considering the correlation between BMI and gender, gender was used as a covariate in the partial correlation analysis. The partial correlation analysis found that BMI was only weakly correlated with MT at a CDG value of 4.32 and weakly correlated with ER at CDG values of 12.97 and 21.62. BMI was not significantly correlated with EOF and was weakly correlated with the slope of the MPF for ECU and FDS at a CDG value of 2.16 (see Table 6).

Table 5.

Results of the Spearman correlation analysis of BMI with MT, ER, EOF, MPF slope and Borg rating scale at different CDG values.

User performance				MPF slope				Borg Fatigue Scale
MT	2.16	n	−0.25	ED	2.16	n	0.27	ED	2.16	n	0.15
	2.16	p	0.18		2.16	p	0.15		2.16	p	0.43
	4.32	n	−0.54*		4.32	n	−0.14		4.32	n	0.06
	4.32	p	0.00		4.32	p	0.47		4.32	p	0.76
	8.64	n	−0.51*		8.64	n	0.00		8.64	n	0.03
	8.64	p	0.00		8.64	p	0.98		8.64	p	0.86
	12.97	n	−0.56*		12.97	n	0.02		12.97	n	−0.02
	12.97	p	0.00		12.97	p	0.92		12.97	p	0.91
	21.62	n	−0.64*		21.62	n	−0.16		21.62	n	−0.06
	21.62	p	0.00		21.62	p	0.41		21.62	p	0.74
ER	2.16	n	0.29	ECU	2.16	n	0.34	ECU	2.16	n	0.24
	2.16	p	0.12		2.16	p	0.07		2.16	p	0.20
	4.32	n	0.44*		4.32	n	0.11		4.32	n	0.13
	4.32	p	0.01		4.32	p	0.55		4.32	p	0.48
	8.64	n	0.56*		8.64	n	−0.12		8.64	n	0.01
	8.64	p	0.00		8.64	p	0.51		8.64	p	0.95
	12.97	n	0.55*		12.97	n	0.18		12.97	n	−0.09
	12.97	p	0.00		12.97	p	0.34		12.97	p	0.65
	21.62	n	0.53*		21.62	n	0.18		21.62	n	0.12
	21.62	p	0.00		21.62	p	0.34		21.62	p	0.52
EOF	2.16	n	0.26	FDS	2.16	n	0.34	FDS	2.16	n	0.10
	2.16	p	0.16		2.16	p	0.07		2.16	p	0.60
	4.32	n	0.50*		4.32	n	0.00		4.32	n	0.08
	4.32	p	0.01		4.32	p	1.00		4.32	p	0.69
	8.64	n	0.33		8.64	n	0.15		8.64	n	0.21
	8.64	p	0.07		8.64	p	0.43		8.64	p	0.27
	12.97	n	0.40*		12.97	n	0.05		12.97	n	−0.04
	12.97	p	0.03		12.97	p	0.80		12.97	p	0.85
	21.62	n	0.37*		21.62	n	−0.05		21.62	n	0.05
	21.62	p	0.04		21.62	p	0.80		21.62	p	0.81

Note. MPF, mean power frequency; MT, movement times; ER, error rate; EOF, effective operation frequency; ED, extensor digitorum; ECU, extensor carpi ulnaris; FDS, flexor digitorum superficialis.

Table 6.

Results of partial correlation analysis of BMI with MT, ER, EOF, MPF slopes and Borg rating scale with gender as the covariate.

User performance				MPF slope				Borg Fatigue Scale
MT	2.16	n	−0.32	ED	2.16	n	0.24	ED	2.16	n	−0.10
	2.16	p	0.09		2.16	p	0.20		2.16	p	0.60
	4.32	n	−0.38*		4.32	n	−0.13		4.32	n	−0.12
	4.32	p	0.04		4.32	p	0.51		4.32	p	0.55
	8.64	n	−0.29		8.64	n	−0.34		8.64	n	−0.02
	8.64	p	0.12		8.64	p	0.07		8.64	p	0.92
	12.97	n	−0.3		12.97	n	−0.26		12.97	n	0.12
	12.97	p	0.12		12.97	p	0.17		12.97	p	0.52
	21.62	n	−0.22		21.62	n	0.00		21.62	n	0.00
	21.62	p	0.26		21.62	p	0.98		21.62	p	1.00
ER	2.16	n	0.12	ECU	2.16	n	0.50*	ECU	2.16	n	0.02
	2.16	p	0.54		2.16	p	0.01		2.16	p	0.94
	4.32	n	0.26		4.32	n	0.08		4.32	n	0.12
	4.32	p	0.17		4.32	p	0.66		4.32	p	0.53
	8.64	n	0.33		8.64	n	−0.13		8.64	n	−0.14
	8.64	p	0.08		8.64	p	0.50		8.64	p	0.46
	12.97	n	0.39*		12.97	n	0.03		12.97	n	−0.24
	12.97	p	0.04		12.97	p	0.88		12.97	p	0.21
	21.62	n	0.42*		21.62	n	0.04		21.62	n	0.12
	21.62	p	0.02		21.62	p	0.82		21.62	p	0.53
EOF	2.16	n	0.26	FDS	2.16	n	0.50*	FDS	2.16	n	0.02
	2.16	p	0.17		2.16	p	0.01		2.16	p	0.92
	4.32	n	0.32		4.32	n	−0.08		4.32	n	−0.12
	4.32	p	0.09		4.32	p	0.68		4.32	p	0.52
	8.64	n	0.18		8.64	n	−0.11		8.64	n	−0.08
	8.64	p	0.34		8.64	p	0.56		8.64	p	0.68
	12.97	n	0.15		12.97	n	0.02		12.97	n	−0.14
	12.97	p	0.45		12.97	p	0.92		12.97	p	0.48
	21.62	n	0.03		21.62	n	0.21		21.62	n	−0.06
	21.62	p	0.87		21.62	p	0.27		21.62	p	0.74

Note. MPF, mean power frequency; MT, movement times; ER, error rate; EOF, effective operation frequency; ED, extensor digitorum; ECU, extensor carpi ulnaris; FDS, flexor digitorum superficialis.

Discussion

The aim of this study was to explore the effects of different CDG settings on muscle fatigue, user performance, and subjective muscle fatigue when engaging in computer-intensive work in an office environment.

The experimental results (see Figure 4) indicate that higher CDG settings do not necessarily lead to better user performance. When the CDG value reaches a certain threshold, a high CDG value may impede the subject's ability to control the cursor position. This result may be attributed to the following factors: 1) Any minor hand movements are magnified by the CDG setting and mapped onto cursor movement; 2) As the CDG value increases, the subject must manipulate the mouse with greater precision. A larger CDG value significantly increases the likelihood of error, which consequently reduces the overall user performance. Hence, it is inadvisable to set high CDG values since accuracy and efficiency will be negatively affected.

The MPF slope distributions showed a significant increase in muscle fatigue when the CDG was set to the most commonly used value (see Figure 5). The experimental data indicates that when the CDG value was greater than 12.97 or less than 8.64, there was less fatigue in both ED and FDS. Notably, at a CDG value of 21.62, not only was there a reduction in fatigue in both ED and FDS, but the overall EOF loss was found to be insignificant, which was similar to the distribution observed for a CDG value of 8.64. To reduce the accumulation of forearm muscle fatigue, it is advisable to avoid setting CDG values between 8.64 and 12.97. This result is similar to a previous study on MT, which concluded that the optimal CDG settings should be 2.4 or 14.5 based on different target distances. The study recommends that CDG settings between 2.4 and 14.5 should be avoided..⁴³

In addition, the subjective fatigue data showed that the subjects generally perceived the most fatigue in their forearm muscles when CDG was set at 8.64 (see Figure 6), while the sEMG data showed that CDG values set between 8.64 and 12.97 were the most likely to cause fatigue. These two findings suggest that there is some consistency between the perceptions of subjective fatigue and the objective data. Moreover, a correlation between subjective fatigue and MPF slope was found by Spearman correlation analysis (n = -0.80, p ≤ 0.05). However, it should be noted that although both the subjective report data and the sEMG data indicated that fatigue was relatively significant (p < 0.05) at a CDG value of 8.64, the subjective report concluded that ECU fatigue was significantly affected by the CDG value setting (p < 0.05), whereas the sEMG data showed that ED muscle fatigue was significantly affected by a CDG value setting (p < 0.05). Although the correlation between subjective and objective data for combined fatigue is consistent with previous findings,^42,44,45 some discrepancy exists between the subjective and objective data regarding the perception of fatigue. This discrepancy may be attributed to the discrepancy between the subjects’ perception of the location of fatigue and the actual location of fatigue when the number of muscles is increased, making the muscles more densely distributed. Therefore, caution should be exercised when using subjective data of a user's perception of fatigue in different muscles for similar studies involving denser muscle distribution.

Optimal user performance was reached at a CDG setting of 12.97 when the highest speed of operation was obtained. However, there was a small increase in error rate compared to lower CDG values (see Figure 4). To maximize efficiency, it is suitable to keep the default CDG value of 12.97. However, for tasks where accuracy is the main requirement, it is recommended that the CDG value should be set at 4.32. Reducing the CDG value from 4.32 to 2.16 resulted in a large marginal increase in accuracy, as the error rate did not change significantly, but there was a significant increase in MT (p < 0.05).

The analysis of gender difference shows that the performance was relatively uniform across genders (see Figure 7). There were no significant differences in MT and EOF when the CDG value was at 2.16, which may be due to the fact that setting the CDG value at 2.16 is an extreme case, so the higher operating difficulty smoothed out the differences between genders. This result further suggests that the CDG value should not be set too low. When the CDG value was greater than 2.16, it was found that males operated faster and with higher effective operating efficiency. Interestingly, for all CDG values, the ER of females was always significantly lower than that of males, and the ER distribution did not change for different CDG values. The ER of males, however, increased significantly as CDG values increased, with the mean ER at the highest CDG setting reaching twice as high as at the ER at the lowest CDG setting. This interesting phenomenon reveals that the distribution of correct rates for the pointer localization task exhibited completely different patterns across genders. Previous research studies may have revealed the reason for this phenomenon. Those studies concluded that women seem to be more risk-averse than men.^46,47 Thus, when the CDG value was changed, women tended to execute the task while maintaining the original error rate, which could mean that women generally invest more effort at higher CDG settings.

The gender difference analysis of MPF slopes shows that females were more susceptible than males for ED and FDS fatigue at specific CDG values, while males were more susceptible than females to ECU fatigue at lower CDG values (see Figure 8). At a CDG value of 4.32, females indicated a a significant decrease in in both ED and FDS fatigue and there was no significant increase in ECU fatigue. At a CDG value of 2.16, there was a significant decrease in males for both ED and ECU fatigue with no significant increase in FDS fatigue. To reduce the risk of upper limb muscle fatigue, it is more suitable for females to set CDG values around 4.32, while it is more suitable for males to set CDG values around 2.16. Males appear to be better adapted to lower CDG settings than females, which could be attributed to the fact that lower CDG settings result in a greater range of arm movement and that testosterone levels are usually higher in men than in women⁴⁸; therefore, male muscles are usually more developed and more easily adapted to a greater range of muscle activity.

In addition, the subjective muscle fatigue data showed that males reported higher levels of fatigue than females, which may be due to the fact that males operate more quickly than females and therefore perform more operational tasks in the same amount of time (see Figure 9).

After excluding the effect of the covariate (gender), correlation analysis found that BMI was only weakly correlated with MT and ER at specific CDG settings and only weakly correlated with the MPF slopes of ECU and FDS at a CDG value of 2.16 (see Table 6). Based on the results of the analysis, BMI is not an important factor in the selection of mouse CDG settings; however, this finding may be due to insufficient sample size. Sample size calculations using G*power software showed that a minimum of 84 subjects were needed to validate the correlation of BMI with the MT, ER, and MPF slopes under specific CDG settings. Future work could expand the sample size to further explore the effect of BMI on mouse design.

The limitations of this study should be considered when interpreting the findings. The results of this study may only be applicable to the specific mouse used in the experiment. Since no studies have yielded results that demonstrate the effect of using a wired mouse versus a wireless mouse on muscle fatigue accumulation, the results of this study only apply to using a wireless mouse. The differences in muscle activity observed in our study were relatively small, possibly due to the short duration of the tasks. In similar muscle fatigue studies, the duration of fatigue experiments was typically 30–120 min,^22,23,33 which is in line with the duration that was set in this experiment. However, the commonly used duration of the fatigue experiment may not have been sufficient to induce pronounced muscle fatigue in the subjects, given the light load of the experimental task. Moreover, the results of this study may be limited due to the selection of a convenience sample. This study chose master's or doctoral students of both genders, aged 22–31 years old, on university campuses because the work environment and job descriptions of this group are more similar to those of office workers. However, this sample cannot be fully equated with office workers, especially since the participants in this study were usually younger than typical office workers. Therefore, the differences between the two groups, such as daily working hours, number of consecutive working days, and physical conditions, should be considered when assessing the limitations of the results.

This study contributes to the field of knowledge by providing guidance on the selection of computer mouse CDG settings to reduce muscle fatigue and fatigue accumulation in office workers who use computer mice. Young office workers with no history of musculoskeletal disorders were recruited for this study and a choice of mouse control-display-gain options was provided for the subjects. Future studies will broaden the scope of muscle examination and test the possibility of reducing muscle fatigue by optimizing CDG settings. In addition to the ED, ECU, and FDS muscles that were investigated in this study, other relevant muscles, such as extensor pollicis brevis, flexor carpi ulnaris, extensor carpi radialis longus, triceps lateral head, biceps, middle deltoid, anterior deltoid and upper trapezius, will be given more attention in subsequent studies.⁴⁹

Although a wireless mouse was used in this study to minimize the influence of extraneous variables, this does limit the generalizability of the findings to the specific context of wireless mouse use. It remains uncertain whether the same outcomes would be observed if a wired mouse was used. Consequently, future research will be conducted to investigate the impact of muscle fatigue at various CDG settings with other wireless devices and wired mice. Such a study could be conducted by examining whether the level of muscle fatigue produced by a wired mouse is similar to that produced by a wireless mouse at various CDG settings for the same tasks and duration.

Conclusion

The results of this study indicate that office workers can reduce their risk of muscle fatigue by adjusting the CDG settings of their computer mice. To reduce forearm muscle fatigue, it is recommended to avoid setting CDG values between 8.64 and 12.97. Office workers can adjust the mouse DPI or the mouse speed setting in the computer system. Upper limb muscle fatigue at work can be reduced by appropriately lowering or increasing the CDG settings. Furthermore, subjective data on muscle fatigue can provide useful information for the field of muscle fatigue measurement, but consideration needs to be given to the accuracy of subjective judgements on the location of muscle fatigue. These findings have practical implications for the prevention of upper limb fatigue and musculoskeletal disorders in office workers. Through discussions with the subjects prior to the study, it was learned that the vast majority of the subjects have not adjusted the CDG of their mouse and have only used the default values set by the computer mouse manufacturer. Consequently, in addition to recommending that office workers adjust the CDG of their mice, it is also recommended that mouse manufacturers change the factory default CDG settings to help reduce the risk of muscle fatigue in office workers.

Footnotes

Abbreviations (sorted by order of appearance):

Acknowledgments

The authors thank all participants who participated in the study.

ORCID iDs

Yihui Ren

Bingchen Gou

Ruiqi Chen

Yihan Gan

Mengcheng Wang

Ao Jiang

Jinlei Shi

Hao Fan

Ethical considerations

The study was conducted in accordance with the Declaration of Helsinki and was approved by the Northwestern Polytechnical University (IRB No. are 201502024 and 202002024).

Funding

This study is partly supported by the Major Project of the National Social Science Fund (NSSF) of China (Grant No. 21ZD11), and the Fundamental Research Funds for the Central Universities (SEU Grant No. 2242024S30013). Additional support comes from the Northwestern Polytechnical University.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Informed consent

All participants were informed about the study and a signed informed consent form was obtained from the individuals who volunteered to participate in the study.

References

Alsobhi

Aldhabi

. Understanding the pattern of musculoskeletal pain and its contributing factors among Hajj pilgrims. Work 2024; 79: 417–424.

Mokhtarinia

Alimohammadi

Sadeghi-Yarandi

, et al. Investigating the relationship between physical, cognitive, and environmental factors of ergonomics with the prevalence of musculoskeletal disorders: a case study in a car-parts manufacturing industry. Work 2024; 79: 215–229.

Mirzaei

Kouhnavard

Daneshmandi

, et al. Determining the effect of environmental factors, work postures and mental workload on musculoskeletal disorders in rescue workers and accidents of one of the water and sewage companies in Iran. Work 2024; 78: 969–981.

Unuvar

Gercek

Tufekci

, et al. The relationship between lower extremity muscle tightness and pain and disability in individuals with non-specific chronic low back pain. Work 2024; 79: 323–330.

Jovanović

Šimunič

. Effect of ergonomic armrest® forearm support on wrist posture related to carpal tunnel pressure during computer mouse work. Int J Ind Ergon 2021; 86: 103220.

Zhao

, et al. Comparing hand gestures and a gamepad interface for locomotion in virtual environments. Int J Hum-Comput Stud 2022; 166: 102868.

Ortega

Williams

Tarre

, et al. 3d Travel comparison study between multi-touch and GamePad. Int J Hum-Comput Interact 2020; 36: 1699–1713.

Chakraborty

Yadav

. Applicability of fitts’ law to interaction with touchscreen: review of experimental results. Theor Issues Ergon Sci 2023; 24: 532–546.

Korek

, et al. Investigating pilots’ operational behaviours while interacting with different types of inceptors 2022: 314–25.

10.

Hemati

Mirjalili

Ghasemi

, et al. Functional parameters, wrist posture deviations and comfort: a comparison between a computer mouse and a touch pen as input devices. Work 2020; 65: 701–706.

11.

Gaudez

Cail

. Effects of mouse slant and desktop position on muscular and postural stresses, subject preference and performance in women aged 18-40 years. Ergonomics 2016; 59: 1473–1486.

12.

Oveisi

Khanlari

Ebrahimzadeh-Afrouzi

. Carpal tunnel syndrome severity, hand discomfort, and usability among three types of computer mouse. Work 2024; 77: 873–882.

13.

Shiri

Falah-Hassani

. Computer use and carpal tunnel syndrome: a meta-analysis. J Neurol Sci 2015; 349: 15–19.

14.

Ring

. Clinical faceoff: routine electrodiagnostic testing is not helpful in the management of carpal tunnel syndrome. Clinical Orthopaedics & Related Research 2016; 474: 1770–1774.

15.

Pattath

Webb

. Computer-usage and associated musculoskeletal discomfort in college students. Work 2022; 73: 327–334.

16.

Keir

Bach

Rempel

. Effects of computer mouse design and task on carpal tunnel pressure. Ergonomics 1999; 42: 1350–1360.

17.

Fagarasanu

Kumar

. Carpal tunnel syndrome due to keyboarding and mouse tasks: a review. Int J Ind Ergon 2003; 31: 119–136.

18.

Szeto

GPY

Sham

KSW

. The effects of angled positions of computer display screen on muscle activities of the neck–shoulder stabilizers. Int J Ind Ergon 2008; 38: 9–17.

19.

Choobineh

Motamedzade

Kazemi

, et al. The impact of ergonomics intervention on psychosocial factors and musculoskeletal symptoms among office workers. Int J Ind Ergon 2011; 41: 671–676.

20.

Skelly

. Assessment of computer workstations for compliance with ergonomic guidelines: a field study. Work 2021; 69: 1019–1026.

21.

Jellinek

Card

. Powermice and user performance. In: ACM; 1990: 213–20.

22.

Wang

Arippa

, et al. Professional and high-level gamers: differences in performance, muscle activity, and hand kinematics for different mice. Int J Hum-Comput Interact 2022; 38: 691–706.

23.

McKinnon

Martel

Callaghan

. The impact of a progressive sit-stand rotation exposure duration on low back posture, muscle activation, and pain development. Ergonomics 2021; 64: 502–511.

24.

Casiez

Vogel

Balakrishnan

, et al. The impact of control-display gain on user performance in pointing tasks. Hum Comput Interact 2008; 23: 215–250.

25.

Penumudi

Kuppam

Kim

, et al. The effects of target location on musculoskeletal load, task performance, and subjective discomfort during virtual reality interactions. Appl Ergon 2020; 84: 103010.

26.

Chen

Lin

. Effects of control-display gain and postural control method on distal pointing performance. Int J Ind Ergon 2019; 72: 45–53.

27.

Kota

Young

. Effects of control-display mapping and spatially dependent gain on supported free-hand gesture pointing performance. Proceedings of the Human Factors and Ergonomics Society Annual Meeting 2019; 63: 391–395.

28.

Agarabi

Bonato

De Luca

. A sEMG-based method for assessing the design of computer mice. Conf Proc IEEE Eng Med Biol Soc 2004; 2004: 2450–2453.

29.

Lin

Chen

Choi

. The postural and control-display gain effects of distal pointing on upper extremity fatigue. Ergonomics 2016; 59: 73–84.

30.

Conroy

Toth

Campbell

. The effect of computer mouse mass on target acquisition performance among action video gamers. Appl Ergon 2022; 99: 103637.

31.

Quemelo

PRV

Vieira

. Biomechanics and performance when using a standard and a vertical computer mouse. Ergonomics 2013; 56: 1336–1344.

32.

Coelho

Lourenço

. Dynamics of forearm muscle activity in slanted computer mice use. Work 2021; 68: 123–135.

33.

Forman

Melchiorre

Holmes

MWR

. Impact of repetitive mouse clicking on forearm muscle fatigue and mouse aiming performance. Appl Ergon 2024; 118: 104284.

34.

MacKenzie

. Fitts’ law as a research and design tool in human-computer interaction. Hum Comput Interact 1992; 7: 91–139.

35.

Xie

Zhou

. Fitts’ law on the flight deck: evaluating touchscreens for aircraft tasks in actual flight scenarios. Ergonomics 2023; 66: 506–523.

36.

Sarker

Norasi

Koenig

, et al. Effects of break scheduling strategies on subjective and objective measures of neck and shoulder muscle fatigue in asymptomatic adults performing a standing task requiring static neck flexion. Appl Ergon 2021; 92: 103311.

37.

Roynarin

Channak

Janwantanakul

. Postural shifts and body perceived discomfort during 1-h sitting when leaning and sitting on an air-filled seat cushion among healthy office workers. Ergonomics 2024; 67: 2126–2137.

38.

Sako

Sugiura

Tanoue

, et al. Electromyographic analysis of relevant muscle groups during completion of computer tasks using different computer mouse positions. Int J Occup Saf Ergon 2017; 23: 267–273.

39.

Bodin

Berglund

Forsman

. Activity in neck-shoulder and lower arm muscles during computer and smartphone work. Int J Ind Ergon 2019; 74: 102870.

40.

Wang

Fan

, et al. Effects of variations in the tragus expansion angle on physical comfort for in-ear wearables. Ergonomics 2022; 65: 1352–1372.

41.

Lin

Hong

Huang

. Influence of virtual keyboard design and usage posture on typing performance and muscle activity during tablet interaction. Ergonomics 2020; 63: 1312–1328.

42.

Hummel

Laubli

Pozzo

, et al. Relationship between perceived exertion and mean power frequency of the EMG signal from the upper trapezius muscle during isometric shoulder elevation. Eur J Appl Physiol 2005; 95: 321–326.

43.

Pang

Hoffmann

Goonetilleke

. Effects of gain and index of difficulty on mouse movement time and fitts’ law. IEEE Trans Hum Mach Syst 2019; 49: 684–691.

44.

Dedering

Nemeth

Harms-Ringdahl

. Correlation between electromyographic spectral changes and subjective assessment of lumbar muscle fatigue in subjects without pain from the lower back. Clin Biomech (Bristol, Avon) 1999; 14: 103–111.

45.

Oberg

Sandsjo

Kadefors

. Subjective and objective evaluation of shoulder muscle fatigue. Ergonomics 1994; 37: 1323–1333.

46.

Charness

Gneezy

. Strong evidence for gender differences in risk taking. J Econ Behav Organ 2012; 83: 50–58.

47.

Gerdes

Gränsmark

. Strategic behavior across gender: a comparison of female and male expert chess players. Labour Econ 2010; 17: 766–775.

48.

Hilton

Lundberg

. Transgender women in the female category of sport: perspectives on testosterone suppression and performance advantage. Sports Med 2021; 51: 199–214.

49.

Wang

Arippa

, et al. The effects of mouse weight and connection type on performance, muscle activity, and preferences among professional gamers. Int J Ind Ergon 2023; 97: 103493.