Force perceptual bias caused by muscle activity in unimanual steering.

This study sought to investigate whether force perceptual bias was affected by differences in posture while steering an automobile using a psychophysical experiment to examine the relationship with muscle activity. The human perceptual characteristics of weight and force are known to be nonlinear, and a perceptual bias can occur, that is, bias that causes a perception of something that is larger or smaller than the actual scale. This is considered to be caused by physical and/or psychological conditions. Sense of effort is believed to be one influential factor. It is known to correlate with muscle activity intensity, and bias may be caused by muscle activity changes. In the current study, we hypothesized that force perceptual bias would depend on posture due to the intensity of muscle activity changes caused by changing postures during steering operation. By investigating this hypothesis, we can clarify the relationship between sense of effort and muscle activity. To investigate this issue, we conducted a psychophysical experiment to confirm postural dependence, and estimated muscle activity using a three-dimensional musculoskeletal model simulation with postural and arm force data during the experiment. In addition, prediction of bias was conducted based on a simulation in the psychophysical experiment using these data. The results revealed that bias existed, as measured by differences in postures. Additionally, a significant moderate correlation was found between the predicted bias and the actual bias, indicating the existence of a relationship between muscle activity and bias.

Introduction

The accurate performance of human movement involves the ability to sense force/heaviness. Human perceptual characteristics are known to be nonlinear; that is, there are differences between actual force and perceived force [1, 2]. Perceived force/heaviness has traditionally been believed to depend on physical (e.g., colors, and surface condition of lifted objects) and/or psychological (e.g., fatigue of muscle) factors, as reported by Jones et al. [3]. De Camp [4] demonstrated that perceived weight is affected by object’s color, reporting that darker-colored objects are perceived to weigh less than lighter-colored objects. Additionally, it is well known that fatigue affects sense of force/heaviness [5-7].

In daily life, an automobile is an example of a system involving a human-machine interaction based on sense of force. Sense of force is thought to be important when driving an automobile, and perceived force changes while driving. Newberry et al. [8] found that the sensation of the force exerted by the steering wheel increases with a power of 1.39, according to Stevens’ power law [9] for steering wheel reaction forces ranging from 5.25–21 N and power of 0.93 for a steering wheel angles ranging from 4–16°. These parameters of power represent the ratio of the intensity of the subject’s perceived exertion of force to the actual exertion. Takemura et al. [10] investigated the perceived force characteristics for a wide range of steering angles using psychophysical experiments and reported that the characteristics followed Weber-Fechner’s law [11]. This law states that perceptual intensity is proportional to the logarithm of the stimulus. It has also been reported that muscle activity changes according to the steering posture of the automobile, which changes sense of force [12].

To investigate perceptual bias, which creates a perception that is larger or smaller than the actual scale, psychophysical experiments have been conducted. Using a psychophysical experiment, van Polanen et al. [13] revealed that bias that overestimates actual weight occurs when there is a visual delay in lifting an object in a virtual reality environment. They investigated the multisensory effect (lifting an object with a visual delay) on the perceived weight [13]. Flanagan et al. [14] found that when lifting an object using a precision grip with the distal pads of the thumb and index finger, bias changed depending on the object’s surface texture. When the surface texture of the lifted object is smooth, the perceived weight increase. Additionally, Sakajiri et al. [15] report that perceptual bias is generated by a difference in the reaction force direction while steering an automobile. Flanagan et al. and Sakajiri et al. report that regarding sense of effort, bias can be affected by whether muscle force functionally acts on movement. This indicates that muscle is one key factor of force/weight perception.

It has previously been reported that sense of effort and perception of force/heaviness are linked because during muscle fatigue or paralysis, we perceive both a sense of increased force/heaviness and an increase in effort [16-19]. Sense of effort is a motor command generated by the central nervous system, and it refers to a signal sent from the brain to a peripheral system. The larger the motor command, the more power a human can exert, and the size of the motor command relates to the sense of effort size. Cafarelli et al. [20] used the intensity of muscle activity as a sense of effort to investigate the relationship between muscle length and sense of force. Moreover, Morree et al. [21] provide neurophysiological evidence that movement-related cortical potential amplitude is correlated with sense of effort. Thus, previous studies have indicated that sense of force/heaviness can be evaluated based on muscle activity, which can be interpreted as sense of effort. The findings described above suggest that bias could potentially be caused by changes in muscle activity with changing postures. However, no previous studies have investigated changes in force perceptual bias caused by changes in postures. It may be possible to explain the generation mechanisms of postural dependence of force perceptual bias by comparing muscle activity intensity, which reflects sense of effort. To investigate this issue, we conducted a psychophysical experiment to confirm postural dependence, and estimated muscle activity using a three-dimensional musculoskeletal model simulation with postural and arm force data during the experiment. Additionally, the prediction of bias was carried out by the simulation in the psychophysical experiment using these data. Overall, in this study, we attempt to clarify human force discrimination in the experiment based on differences in muscle activity.

Materials and methods

Participants

The participants included nine healthy, right-handed subjects (nine males; mean [SD]: 21.8 [1.6] years old; 1.75 [0.06] m; 69.5 [6.7] kg). Of the nine participants, eight have official driver’s licenses, and two drove on a regular basis. All participants gave written informed consent before participating in the study. Participants were paid for their time. The experimental procedures were previously approved by the local research ethics committee (Nagoya Institute of Technology).

Apparatus

We used a simulated steering device in the experiment, as shown in Fig 1. It was the system used in [15]. A six-axis force sensor (BL Autotec, Ltd., Micro 5/50-S09) was attached at the base of each handgrip to obtain the exerted force from the hand, and the torque presentation was generated by two servomotors (maxon motor, RE40) attached to one end of the main driveshaft. Each servomotor was attached to a 21:1 reduction gear (Harmonic Drive Systems Inc., HPG-14A-21) and a rotary encoder (Microtech Laboratory Inc., ME-20) in order to apply the desired reaction force and obtain the angle. The curved handgrips were made of acrylic plastic and formed two arcs of a circle 350 millimeters in diameter.

Fig 1

The simulated steering wheel device.

The device included two servo motors to generate steering wheel torque and a six-axis force sensor was attached at the base of each handgrip to obtain the exerted force from the hand. However, we only used the right side grip in this experiment. It was the system used in [15].

Procedure

Psychophysical experiments were performed using the staircase method, which included downward step and upward step, in which the test stimuli deviates from the reference stimulus(very large and very small, respectively). In this case, very large/small means the subjects could definitely perceive the difference from the reference stimulus. These test stimuli were confirmed before the experiment. The subjects were asked to compare the magnitude of the reaction force in the reference posture and the experimental posture. They grasped the handgrip with the right hand only. Each experimental posture is shown in Fig 2. The reference posture was the initial position of the steering (0°), and the experimental postures were static postures of 30°, 60°, −30° and −60° from the reference. The reference stimulus was 2.0 Nm, and the experimental stimuli were changed in ascending or descending stepwise increments of 0.2 Nm between 1.1–2.9 Nm. The experimental postures and magnitude of each test stimulus were decided from the realistic condition [12]. The direction of the force was the same between the reference and the test stimuli. The steering wheel rotated to the left at the experimental postures of 30° and 60°, and the steering wheel rotated to the right at −30° and −60° because the direction of the steering reaction force was the same as that of the actual steering reaction force. The experimental tasks were as follows.

Fig 2

Experimental conditions.

The subject grasped the right side of the steering wheel device and memorized the reference stimulus in the reference posture(0°). Afterward, the posture changed to the experimental postures(30°, 60°, −30, −60°) and memorized the test stimuli (1.1–2.9 Nm, staircase method including upward step and downward step). Then, the subject was asked about the larger stimulus. The tasks were repeated 50 times(25 upward step and 25 downward steps) in each experimental posture.

The participant grasped the handgrip with the right hand and memorized the magnitude of the reference stimulus presented in the reference position for 3 seconds. The participant maintained the posture while the stimulus was presented.

After changing to the experimental posture, the participant memorized the magnitude of the test stimulus presented for 3 seconds. The participant maintained the experimental posture while the stimulus was presented.

The participant was asked which side was larger.

The subsequent test stimulus was modified based on the participant’s response.

According to the response of each trial, the test stimulus of the next trial for downward step and upward step was changed as follows.

Answer that the test stimulus was larger than the reference stimulus: reduce the test stimulus by 0.2 Nm.

Answer that the test stimulus was smaller than the test stimulus: increase the test stimulus by 0.2 Nm.

These procedures were used in both downward and upward step. The downward and upward step were conducted alternately. The test stimulus was repeated at the comparative stimulus of the chance level that is, near the subjective equivalence value. To avoid the effect of fatigue, a break was provided for each posture. To avoid the order effect, the order of the experimental posture was randomized for each participant. A complete experimental session for each participant consisted of 200 steering trials, with 25 upward and 25 downward steps in each posture.

Muscle activity estimation using a 3D musculoskeletal model

We used OpenSim [22] to calculate the muscle activity in each experimental condition. Muscle strength was calculated using a combination of elastic and contractile elements based on Hill’s muscle model reported by Thelen [23]. The muscle parameters, such as the maximum isometric muscle strength F, optimum muscle fiber length l, and pennation angle of the muscle, were determined according to a previous study [24]. In the muscle activity calculation, we measured the reference posture (0°) and the experimental posture (30, 60, −30, −60 °) using six motion capture systems (Optitrack, Optitrack Flex3), and joint angle and joint torque were calculated using inverse kinematics and inverse dynamics. The reflex marker was attached to the shoulder, elbow, wrist, and hand, as shown in Fig 2. Muscle strength was determined by optimizing the muscle activity to balance the joint torque. The m-th muscle was calculated to satisfy the following Eq 1. is the isometric maximum muscle strength, τ is the joint torque at the j-th joint, and r is the moment arm. α represents muscle activity and is a continuous function of α(0 ≤ α ≤ 1), which can be regarded as a control signal in the musculoskeletal system [22]. Based on the relationship between the motor unit firing frequency and muscle activity, the higher the motor unit firing frequency, the greater the muscle activity [25]. The moment arm was determined by the m-th muscle length l and the j-th joint angle [26, 27]. The following shows the relationship between muscle strength F and muscle activity α. is the normalized muscle fiber length, and is the normalized muscle strength-length relationship. We used the parameter of and from a previous study [24].

Data analysis

Perceptual bias

In the psychophysical experiment, we calculated the perceptual bias to determine whether a perceived force with an experimental posture was perceived differently when compared with a reference posture. The percentage of responses indicating that the test stimulus was “larger” were calculated for each presented comparison. The percentages were plotted, and a psychometric curve was fitted to the points with a cumulative Gaussian distribution: where μ and σ are the fitted parameters representing the mean and SD of the curve, respectively. Because some experimental stimuli were presented more often than others, a weighted least squares fit was used [28]. The value of μ represents the perceptual bias and 2.0 + μ represents the points of subjective equality for a specific session. A positive value represents an overestimate (i.e., the reference stimulus was perceived to be larger than it actually was in the experimental posture). In contrast, a negative value represents an underestimate (i.e., the reference stimulus felt lighter in the experimental posture). The average bias for all subjects was calculated from the experimental results, and the comparison was carried out using Student’s t-tests (significance level: 5%) between the reference and experimental postures. Additionally, analysis of variance (ANOVA) was performed between the experimental postures, and pairwise comparisons using the Holm method were performed (significance level: 5%).

Prediction of perceptual bias from muscle activity

In the muscle activity estimation, postural data that were obtained using a motion capture system and force data during each trial were used. In operation of the steering wheel, the previous study reported that the arm and shoulder move to make the positive tangential steering force by moving with forward elevation. For the negative tangential steering force, the arm and shoulder move in a downward direction [29]. These movements are created from the deltoid muscle (anterior, medial, and posterior), the pectoralis major muscle (upper and medial portion), the biceps brachii(long and short), and the triceps brachii (long head and lateral part). Therefore, we used these muscles as representative muscles. In this study, we used the average of the above-mentioned four muscles. The muscle activity differences between the experimental postures were compared using ANOVA, and pairwise comparisons were performed using the Holm method (significance level: 5%). In a previous study, we proposed an estimation model of the force perception-change ratio using muscle activity during steering [12]. It is possible to estimate the perceived force using this model. In this model, the magnitude of perceived force was measured using the psychophysical experiment, and logarithmic fitting was performed based on Weber-Fechner’s law: where F is the perceived force, F is the applied force, and a and b are coefficients obtained using the least square method. In addition, muscle activity against the force magnitude F was estimated using a three-dimensional musculoskeletal model. The muscles used in the muscle activity estimation were the same as those described above. We obtained the linear relationship between the F and the muscle activity. where α is the muscle activity, and k and m are the coefficients obtained using the least-square method. By substituting Eq 6 into Eq 5, we obtained the following equation:

Table 1 lists the coefficient values in Eq 7. This equation can be expressed as a function of the muscle activity. Using Eq 7, the perceived force F can be predicted from the muscle activity.

Table 1

Model coefficients.

Coefficient	Value
a	11.74
b	−14.91
k	0.0015
m	0.00059

We predicted the perceptual bias in each posture using these equations. First, muscle activity was estimated using the stimulus force data and posture (reference and experimental, respectively). The F values were then estimated in both conditions, and a comparison was carried out. In cases where the F of the test stimulus was larger than that of the reference stimulus, we recorded the response as “larger”. The calculation method of the force perceptual bias followed the technique described in the “Perceptual Bias” chapter, and the predicted force perceptual bias μ was calculated. The accuracy was verified by obtaining the correlation coefficient between the true value and the predicted value.

Results

Force perceptual bias

In this experiment, we investigated the perceptual bias in driving posture using a psychophysical experiment. Fig 3 shows the results of the psychophysical experiments on the representative subjects. Fig 3(a) shows the trajectory of a given test stimulus during the experiment. It is predicted that the subject overestimated the reference stimulus because the plots are mostly located at positions larger than 2.0. In Fig 3(b), a psychophysical curve was calculated using the results of Fig 3(a). A positive perceptual bias existed because the center of the “larger steering force”(PSE = 0.5) shifted to greater than the reference stimulus. Fig 4 shows the average of the perceptual bias calculated from the results of the psychophysical experiment. The bias for each posture was compared with the reference posture using Student’s t-tests (significance level: 5%). Significant differences were found at 30° (t = 2.7, p = 0.03), −30° (t = −9.0, p < 0.001), and −60° (t = −6.5, p < 0.001). No significant differences were observed at 60° (t = 0.16, p = 0.9). An ANOVA revealed significant differences (F1,8 = 28.3, p < 0.001) between each experimental posture. In pairwise comparisons, significant differences were observed at 30° versus −30°(t = 9.4, p < 0.001), 30° versus −60°(t = 7.2, p < 0.001), 60° versus −30°(t = 4.7, p = 0.002), and 60° versus −60°(t = 4.1, p = 0.01). No significant differences were observed at 30° versus 60°(t = 2.6, p = 0.06) and −30° versus −60°(t = 1.3, p = 0.2). The results show that perceptual bias existed in each experimental posture except for 60°. Additionally, it was shown that there is a significant difference in the size of the bias based on the posture.

Fig 3

A typical result from the psychophysical experiment.

(a) shows the trajectory of a given test stimulus during the experiment on the representative subject. Each step was given 25 times. In (b), each data point shows the percentage of responses in which the test stimulus was reported as “larger”, calculated for each presented comparison. The solid line shows the psychometric curve fitted to the answer plot with a cumulative Gaussian distribution using the weighted least-squares method.

Fig 4

The result of the psychophysical experiment.

A positive value represents an overestimate (i.e., the reference stimulus was perceived to be larger than it actually was in the experimental posture). A negative value represents an underestimate (i.e., the reference stimulus felt lighter in the experimental posture).

Muscle activity estimation and bias prediction from muscle activity

The psychophysical experiment showed that perceptual biases existed in the experimental postures (except for 60°). To further investigate the perceptual bias, we estimated the muscle activity in the experimental postures. Fig 5 shows the representative results of the muscle activity estimation. As a representative results, 1.9 was chosen because it was found most frequently among all subjects in the experiment. An ANOVA revealed significant differences (F = 1.2 * 103, p < 0.001) between each experimental posture. In pairwise comparisons, significant differences were observed at 30° versus 60°(t = −3.35, p < 0.001), 30° versus −30°(t = −52.2, p < 0.001), 30° versus −60°(t = −27.1, p < 0.001), 60° versus −30°(t = −49.7, p < 0.001), 60° versus −60°(t = −24.3, p < 0.001), and −30° versus −60°(t = 25.0, p < 0.001). Fig 6 shows the muscle activity estimation result of the reference angle. These muscle activities are compared using Welch’s t-tests (significance level: 5%). The result showed a significant difference between the directions of the force(t = −1.93 * 102, p < 0.001). Fig 7 shows the plots between the predicted bias μ and the measured bias. We obtained a significant, moderate correlation (r = 0.56, p = 0.0028). These results indicate that muscle activity varied with posture, suggesting that muscle activity affected differences in the perceptual bias.

Fig 5

Mean muscle activity for each angle in all participants (with a stimulus of 1.9 Nm).

The 1.9 Nm test stimulus was used as a force value in the muscle activity estimation because this stimulus trial was the most common across all participants and postures.

Fig 6

Mean muscle activity of reference angle in all participants.

In 30° and 60°, the anti-gravitational force is given for the reference force in the 0° posture, and the gravitational force is given for the reference force in the 0° posture in −30° and −60°. The Welch’s t-test showed a significant difference between the directions of the force(t = −1.93 * 102, p < 0.001).

Fig 7

The scatter plots show the predicted bias from the calculation and the measured bias from the psychophysical experiment.

The solid line is the line of equality, where the predicted bias and measured bias exactly matched. We obtained a significant moderate correlation (r = 0.56, p = 0.0028).

Discussion

Bias and muscle activity

The results revealed significant differences when compared with the reference posture (i.e., force perceptual bias was caused by changing the posture) in all positions except for 60°. Additionally, the muscle activity estimation was also carried out in each trial. As shown in Fig 5, muscle activity varied depending on the posture, even when the same stimulus was presented to participants. Jones reported that when the weight of an object is discriminated, the relative size is perceived and scaled by the range of muscle activities involved in motion [3, 30]. This finding indicates that high muscle activity could potentially cause perceptual bias.

The results of the psychopsysical experiment revealed a significant difference between postures, as shown in Fig 4. Additionally, significant differences existed between the anti-gravitational and gravitational directions, as shown in Fig 6. These results suggest that the force direction during the trials affected the perceptual characteristics, similar to the results of Sakajiri et al., who reported an effect of whether the force direction was in the gravitational direction or not [15].

Human somatic sensation is known to change depending on whether the direction of the force is in the gravitational direction or not, and many studies have examined the effects of gravity. In the field of developmental psychology, Hood et al. report that infants learn the effect of gravity on objects as they age [31]. People move on the assumption that there is gravity [32], and the weight discrimination threshold rises in zero-gravity space [33]. In addition, Young et al. report that the positional sense of the body is lost, and motor skill decreases, when subjects operate in the absence of vision under zero-gravity space conditions [34]. The direction of the reaction force changes depending on the rotating direction in steering and becomes the anti-gravitational direction depending on the position of the arm. In the 30° and 60° conditions, the force direction is anti-gravitational because only the right hand gripped the steering wheel in this experiment. The reaction force can be offset by the arm’s own weight in the anti-gravitational direction. Therefore, the muscle activity becomes low at 30° and 60°. Perceptual bias would also be expected to be affected by the difference in muscle activity with the direction of force.

Bias prediction

We conducted a psychophysical simulation experiment to predict bias using estimated muscle activity from postures and arm force data during the experimental tasks. The results revealed a significant moderate correlation between the predicted bias and the actual bias, indicating that human force discrimination could be predicted relatively accurately based on the psychophysical experimental simulation. Since only the estimated muscle activity was used, the prediction made it easier to examine the bias, compared with the experimentally determined muscle activity. Additionally, from the perspective of the force perception mechanisms of the body, it is possible to explain the bias based on muscle activity. Consideration of perceptual bias in steering is useful for designing steering reaction force, and the improvement of operability could play an important role in preventing operational error.

In recent years, however, it has been reported that afferent signals from muscle spindles and skin receptors in the periphery are also important factors in determining sense of force/heaviness [19, 35, 36]. Although it has been confirmed that the sense of effort can be used for judging force/heaviness, an influential current hypothesis predicts that judgments of force/heaviness are based not only on sense of effort but also on feedback of afferent signals returning from the periphery [37]. Monjo et al. propose that humans do not perceive signals of only efferent or afferent signals as sense of effort but can perceive effort by changes in weight between both signals according to the experimental conditions [38]. The present experiment did not include conditions such as paralysis of muscle spindles. However, as Proske et al. report, it is necessary to provide participants with proper instructions when examining either efferent or afferent signals alone [19]. Since the prediction is carried out only by the muscle activity interpreted as the efferent signal, the afferent signals, such as sensing information from the muscle spindle and cutaneous sensation, which can be considered afferent feedback, appear to affect the prediction accuracy.

Additionally, although the range of steering reaction force is the same in the estimation model, the model was based on psychophysical experiments using both hands. Therefore, the current model cannot be completely applied in this case.

The accuracy of predicting perceptual bias depends on the accuracy of estimating the muscle activity using the musculoskeletal model. In this estimation, muscle co-contraction is neglected in the estimation of muscle activity using our method. Humans are known to perform stable movements by increasing joint stiffness through muscle co-contraction [39-41]. Therefore, it is important to consider muscle co-contraction when estimating muscle activity, to improve estimation accuracy. Additionally, Osu et al. report that muscle co-contraction decreases as humans become accustomed to motor tasks [42]. In other words, it is possible to improve accuracy by reducing the effect of co-contraction by setting experimental conditions under which co-contraction does not occur, or by selecting subjects who are familiar with such motor tasks.

Conclusion

In the current study, we investigated whether force perceptual bias depends on posture while steering using a psychophysical experiment. The results revealed bias at postural angles of 30°, −30°, and −60°. These findings suggested that muscle activity increases by changing the posture and direction of the reaction force. We predicted the force perceptual bias using muscle activity during the experimental task and obtained a significant moderate correlation between the predicted and measured bias. The results of the prediction indicated that it is possible to predict perceptual bias with relatively high accuracy using muscle activity, interpreted as sense of effort. In future studies, we plan to test steering reaction force conditions considering this perceptual bias, to investigate the relationship with the sensation of steering.

24 Jul 2019

PONE-D-19-16009

Force perceptual bias caused by muscle activity in unimanual steering

PLOS ONE

Dear Mr. Kishishita,

Thank you for submitting your manuscript to PLOS ONE, and I'd like to personally apologize for the delay in handling the manuscript.

After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria (https://journals.plos.org/plosone/s/criteria-for-publication) as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Both reviewers and I agree that more information, and possibly new analyses, are needed to comply with Criterion #4 (Experiments, statistics, and other analyses are performed to a high technical standard and are described in sufficient detail). Please see Reviewer 1's comments regarding experimental details that seem to be missing. Also, some analyses and model equations may not be appropriate, as pointed out by Reviewer 2. I too wonder whether the conclusions would remain the same if torque direction were treated separately from posture per se, whether averaging muscle activity (e.g. across agonist/antagonist) was appropriate, and why EMG measurements were not made.

Additionally, to ensure adequate compliance with Criterion #3 (Conclusions are presented in an appropriate fashion and are supported by the data), please address the comments from both reviewers about the clarity of hypotheses, predictions, and conclusions. The manuscript is admirably concise, but sometimes too much so, leaving the reader with important unanswered questions.

Lastly, it was not entirely clear whether data will be uploaded as part of Supplementary Information, or uploaded to a public repository, as outlined in Criterion #7.

We would appreciate receiving your revised manuscript by Sep 07 2019 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Christopher R. Fetsch

Academic Editor

PLOS ONE

Journal Requirements:

1. When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for stating the following in the Acknowledgments Section of your manuscript:

This work was partially supported JSPS KAKENHI Grant Number JP17J06986, JST, PRESTO Grant Number JPMJPR16D3 and laboratory exchange program via Special Interest Group for Haptics, The Virtual Reality Society of Japan.

We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

JSPS KAKENHI Grant Number JP17J06986

JST, PRESTO Grant Number JPMJPR16D3

3.  We noticed you have some minor occurrence of overlapping text with the following previous publication(s), which needs to be addressed:

https://ieeexplore.ieee.org/document/8341739

In your revision ensure you cite all your sources (including your own works), and quote or rephrase any duplicated text outside the methods section. Further consideration is dependent on these concerns being addressed.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: No

Reviewer #2: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: I Don't Know

Reviewer #2: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The framework of this study is difficult to understand, as the introduction does not provide a clear link between sense of effort and posture. Further, there is no clear statement of hypotheses/predicted outcome, which, given the limited framework of the introduction make it difficult to interpret the purpose of the study. Further, there is no description of the raw data that was collected and only report of a reduced bias term in a very limited results section – this does not give the reader much to work with in terms of interpreting the results of the study. There are many details missing from the methods, including how trials were counterbalanced, other than report that they were not equal. This is worrisome, as having more “heavy” trials can bias subjects to make judgements that a stimuli is more “heavy” and vice versa. A properly counterbalanced study will allow for careful examination of perception without the confound of unequal trial presentation (and a more "true" judgement of whether or not something requires more perceptual effort). Additionally, it appears that they are adopting parameter values in the muscular model that were originally described in studies of 30-70 year old subjects, while this population is ~20 years old.

Major

1) The abstract is difficult to read and needs to clearly portray the motivation of the study.

2) The Introduction is lacking many details to give the reader the necessary information to support the motivation for the study: Line 18-19 – what are the physical and psychological factors impacting sense of effort? What do the numbers reported for Stevens’ power law mean about actual perceived and executed muscular effort? Why isn’t there a description of Weber-Fechner’s law (some readers likely do not know what this is)? How does muscle activity change the sense of force? It would help the reader to have some insight into the physiologic basis of WHY perception of effort occurs and why it is important, especially since their study collects and models EMG data. Overall, the introduction just gives a list of studies, but does not provide any framework with which to understand the motivation of the authors experiments.

3) I would recommend refraining from referring to upper limb control as posture, as most readers in the field of motor control would assume that it refers to lower limb posture and balance.

4) Line 57-59: The authors do not present sufficient evidence that “sense of force and heaviness can be estimated from muscular activity, which can be interpreted as sense of effort”- it is unclear how this statement is true given the evidence that the authors have presented.

5) The authors do not present any information for the specifics of the staircase methodology – what specific steps were used? What happened if subjects made an error? Why were the postures of 30 and 60 degrees chosen? Line 11 – How were experimental stimuli modified based on the participant’s response?

6) This task has a significant memory component, in that subjects have to hold the posture for 3 seconds and memorize it, how do the authors account for aspects of memory that may contribute to the task?

7) Line 3 - Heaviness typically refers to the weight of an object (eg., holding it and testing weight in your hand). The subjects are not supporting the steering wheel, but are instead gripping the wheel in response to a force being applied. It is unclear whether the authors are testing response to heaviness or response to an applied force. As the directions they have supplied to subjects do not match up with the description of the research in the introduction.

8) Line 114-116: The authors describe 200 trials, with 25 repetitions in each mode, but there is no description of what these modes were. Clarification and explicit statements are experimental procedures are needed here.

9) Why wasn’t EMG collected and then confirmed via the model-based approach?

10) The utilization of parameters proposed by Thelen is surprising given that this original citation examined age-related changes in biomechanics from age 30-70 (young vs old adults). The sample here is roughly ten years younger and it is likely that these same parameters do not apply to younger adults.

11) The results section of this manuscript is two small paragraphs. This gives the reader very little information to go on in terms of what the outcome of this study is and how it relates to their hypotheses and framework.

Minor

1) Perceptual bias needs to be clearly defined in the abstract

2) 9 subjects seems like far too few for a study like this. Is this study appropriately powered? Why was one of the subjects unable to drive? Does this subject have extensive driving experience?

3) Line 113 – How long were the breaks that subjects were allowed?

4) Why weren’t the experimental stimuli evenly counterbalance. It is possible that presenting with more “heavier” trials will bias subjects to over-estimate, based on previous history of stimuli.

5) Figure legends need more description.

Reviewer #2: The manuscript describes a study using a steering wheel where different forces (torques) are applied on the hand when holding the wheel in different orientations. The authors show that the orientation results in biases of the perceived force. Furthermore, they link this to differences in muscle activations that were calculated from the measured reaction forces using a muscle model.

I think this is an interesting study, relating muscle activity to force perception. However, I have some concerns about the terminology and the data analysis/methods. Although the results are nice, I think some of the analyses are not valid for the current data set and this affects the interpretations. Also, I believe the results as they are now do not only show effects of posture, but also of torque direction. The authors should perform new analyses to address these comments.

1) First of all, the authors report that they investigate effects of force. However, since the task concerns a rotation of the steering wheel, this is actually about torques, not forces. Indeed, the biases are reported in Nm, not N. Although the participants are asked to report which stimulus was heavier, the authors should clarify that they measure torques.

2) One main concern about the methods is that the different postures also involve different torque directions. I see that this was done to have realistic conditions in driving and steering, but this could influence the results. In fact, it seems that the different orientations do not differ in perceptual bias (i.e. no difference between 30 and 60 degrees, neither between -30 and -60 degrees), but the bias direction (i.e. negative or positive) is different. So the biases also seem to be driven by the direction of the torques. By altering the direction, the authors seem not to have not only investigated the effect of posture (with different muscle contributions) on force perception, but also the effect of torque direction (which should also lead to different muscle contributions based on the direction of the force that is applied by the arm) on perception. Since the same torque direction was used in the reference as the experimental condition, this still investigates an effect of posture for each condition on its own. Therefore, I think the results are still interesting, since significant biases were found, but only postures with the same force direction can be compared. In sum, I think the authors should split their analysis by the two torque directions.

3) Related to this, I think the authors should report the muscle activity for the reference condition as well. These were performed with two different torque directions. If the muscle activity differs between these directions in the same posture, this would already show that the muscle activity also varies because of torque direction effects, not only posture.

4) Another concern I have about the methods is the muscle model. To me, it seems strange that the muscle activity of agonist and antagonists are averaged. They mention this as a limitation in the discussion, but I think the muscle model as it is now is not very insightful. The average activity of multiple muscles can give similar values for very different combinations of different muscles. Perhaps they could calculate the activity for different muscles (or at least only average over agonist and antagonist muscles). If the activity pattern is similar in different tasks and only increases overall, this would validate their decision to average the muscle activity, but now this is unknown.

5) It seems to me that the calculations for the predicted bias are unnecessarily complicated. I do not think the perception-change ratio is needed to do this. When equation 5 and 7 are combined: Fp=a log (alpha-m/k)+b. Therefore, equation 6, 8 and 9 do not seem necessary.

6) I do not understand where the values in Table 1 come from. Are they from a previous paper or are they fitted to the data? If the latter is true, what data is exactly compared to which model to obtain these parameter values?

7) The authors correlate the predicted biases from the muscle activity/measured forces to the measured biases. However, I wonder whether they can do it the way they have done it now. First, similarly to my earlier comment about comparisons between the torque directions, it might not be fair to correlate over two different torque directions. The biases in one direction are negative, while the biases in the other direction are positive. I am not sure whether it is justified to combine these two conditions into a single plot. Secondly and related to this, the authors seem to have combined different data points (for different conditions) for the same subjects into one plot. Different conditions for one subject are within factors, whereas data points for different subjects have between factors and therefore have different sources of variation. I am not an expert on these statistics, but I wonder whether this can be combined in a single correlation. Perhaps a multilinear regression model might be more suitable.

8) Do the authors have an explanation why they do not find a bias for the 60 degree condition? This does not follow from the muscle activity pattern, so perhaps the authors could indicate other factors that drive the perception?

Minor comments:

a) P2 L46 I am not sure whether visual information can be considered a cognitive feature. This could also be more low-level sensory effects that contribute to the ultimate heaviness perception.

b) The authors report that they counterbalance the conditions to avoid order effects. However, it seems that the reference condition was always performed before the experimental condition. Could this not also lead to order effects or biases? See for example the time-order error as described by Hellstrom (1985, Psychological Bulletin; 2003 Perception & Psychophysics).

c) P5 L158 The authors state that they test whether the bias was different between the standard and experimental procedure. However, from the psychometrical curve, the bias is defined as the difference between the standard and the experimental stimulus, so there can be no bias calculated for the standard position. Do the authors mean they tested whether the biases were significantly different from zero?

d) P6 L195 I think this should be Eq. 8, not Eq. 1?

e) The author contributions are not described (P 9).

Textual comments:

- P1 L18 "..to sense of force..." remove "of"

- P1 L24 "Thus" seems strange here, since fatigue affects do not relate to the previous sentence about colour affects. Perhaps replace with "In addition".

- P5 L178 "ANOVA was performed" -> An ANOVA was performed

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Jennifer Semrau

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

13 Sep 2019

Thank you for your valuable comments.

We have revised the manuscript according to the comments from editor and reviewers.

All answer are including in attached files.

Thank you in advance.

Respectfully,

Yusuke Kishishita

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

24 Sep 2019

PONE-D-19-16009R1

Force perceptual bias caused by muscle activity in unimanual steering

PLOS ONE

Dear Mr. Kishishita,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised by the Academic Editor. (A point-by-point rebuttal letter is not necessary for this revision, as there are no additional comments by the Reviewers.)

Publication Criterion #5 states: “The article is presented in an intelligible fashion and is written in standard English.” Now that the scientific concerns raised by the reviewers have been addressed, it was easier to discern whether this criterion has been met, and the short answer is no, it has not. The manuscript requires additional proofreading and revision to improve clarity, preferably with the help of an expert in written English. [Fortunately I have done much of this work for you, see below. -CF] We apologize for not raising this concern in the previous round, but it was more critical to get the scientific issues out of the way first.

Note that PLOS ONE does not copyedit manuscripts, and thus what may seem like minor errors that would be handled by editorial staff at other journals must here be rectified by the authors. In general, things to look out for include: missing articles (a, the), verb tense and subject-verb agreement, and pluralization of nouns, and spelling errors. I had planned on highlighting just two or three examples, but in the process I ended up writing down every error I found, so I might as well provide you the whole list (of course there may be others that I missed):

line 12: acitivity -> activity.

12-14: The two sentences “This study enable us…” and “In addition, it can…” are difficult to understand, and seem to relate more to broader implications or extensions of the study rather than the study itself. I would omit them, or move them to Discussion.

38: “These power represent”

Fig 2 Legend: gripped, not griped. Memorized, not memorize. Were, not are.

101-113: several singular/plural errors etc. in this section

132: “the test stimulus is smaller than the test stimulus” — you mean reference?

147: “according to a previously investigated [23].”

189: “In operation of steering wheel, we mostly used the arm and shoulder to make positive tangential steering force by moving forward elevation” — This sentence is problematic because “we…used” implies you are referring to the experimenter, not drivers in general, and thus the reader is expecting the sentence to describe something about the task or apparatus rather than a justification for including particular muscle groups in the muscle activity analysis.

197: “, as the average of the four muscle activities described above” repeats an earlier part of the sentence and can be omitted.

197-199: Sentence about ANOVA+Holm is nearly identical to the one on 184-186.\\

228: semicolon or period instead of comma

231: larger steering force, not larger steering wheel

Fig 3b legend: each data point, not “each plot”

Legends for Figs 4+5: the statistics don’t need to be repeated verbatim from the results text.

238 & 253: “In multiple comparison,“ recommend using the word “pairwise”, e.g. “In [multiple] pairwise comparisons”, and also suggest using “vs.” or “versus” instead of an dash (in for example  “30 — -60”) so it’s easier to read when the angles are negative.

247: “showed that the perceptual bias exist” —> showed that perceptual biases exist, or bias exists

256: “ To evaluate the muscle activity in different direction of force,”  ??

257: “muscle activity estimation result of reference angle” ??? As a result of?

274: “of psychopsysical experiment” — misspelled and missing a “the”

275: “ the significant difference exist”

I found the rest of this subsection of Discussion to read pretty well, so I stopped proofreading at that point. However you may want to check over the remaining sections and double-check the figure legends and axis labels etc. as well.

We would appreciate receiving your revised manuscript by Nov 08 2019 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'

A point-by-point rebuttal letter is not necessary for this revision.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Christopher R. Fetsch

Academic Editor

PLOS ONE

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

30 Sep 2019

Thank you for reviewing my language issue and sorry for bothering your time.

We edited the manuscript and submitted the language editing service.

Please see our manuscript again.

Thank you in advance.

Respectfully,

Yusuke Kishishita

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

2 Oct 2019

Force perceptual bias caused by muscle activity in unimanual steering

PONE-D-19-16009R2

Dear Dr. Kishishita,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Christopher R. Fetsch

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

10 Oct 2019

PONE-D-19-16009R2

Force perceptual bias caused by muscle activity in unimanual steering

Dear Dr. Kishishita:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Christopher R. Fetsch

Academic Editor

PLOS ONE