Limits...
Unintentional Interpersonal Synchronization Represented as a Reciprocal Visuo-Postural Feedback System: A Multivariate Autoregressive Modeling Approach.

Okazaki S, Hirotani M, Koike T, Bosch-Bayard J, Takahashi HK, Hashiguchi M, Sadato N - PLoS ONE (2015)

Bottom Line: As for the causality analysis, noise contribution ratio (NCR), the measure of influence using a multivariate autoregressive model, was also computed to identify the degree to which one's postural sway is explained by that of the other's and how visual information (sighted vs. blindfolded) interacts with paired participants' postural sway.It was found that for synchronization to take place, it is crucial that paired participants be sighted and exert equal influence on one another by simultaneously exchanging visual information.Furthermore, a simulation for the proposed system with a wider range of visual input showed a pattern of results similar to the behavioral results.

View Article: PubMed Central - PubMed

Affiliation: Division of Cerebral Integration, National Institute for Physiological Sciences, Okazaki, Aichi, Japan.

ABSTRACT
People's behaviors synchronize. It is difficult, however, to determine whether synchronized behaviors occur in a mutual direction--two individuals influencing one another--or in one direction--one individual leading the other, and what the underlying mechanism for synchronization is. To answer these questions, we hypothesized a non-leader-follower postural sway synchronization, caused by a reciprocal visuo-postural feedback system operating on pairs of individuals, and tested that hypothesis both experimentally and via simulation. In the behavioral experiment, 22 participant pairs stood face to face either 20 or 70 cm away from each other wearing glasses with or without vision blocking lenses. The existence and direction of visual information exchanged between pairs of participants were systematically manipulated. The time series data for the postural sway of these pairs were recorded and analyzed with cross correlation and causality. Results of cross correlation showed that postural sway of paired participants was synchronized, with a shorter time lag when participant pairs could see one another's head motion than when one of the participants was blindfolded. In addition, there was less of a time lag in the observed synchronization when the distance between participant pairs was smaller. As for the causality analysis, noise contribution ratio (NCR), the measure of influence using a multivariate autoregressive model, was also computed to identify the degree to which one's postural sway is explained by that of the other's and how visual information (sighted vs. blindfolded) interacts with paired participants' postural sway. It was found that for synchronization to take place, it is crucial that paired participants be sighted and exert equal influence on one another by simultaneously exchanging visual information. Furthermore, a simulation for the proposed system with a wider range of visual input showed a pattern of results similar to the behavioral results.

No MeSH data available.


Related in: MedlinePlus

Relation between participants’ influence and time lag for postural sway synchronization.Each dot in the figure represents the average of all trials for each of the paired participants. The data are plotted in the following way. X-axis values represent differences in paired participants’ influence on one another’s postural sway (estimated by ΣNCR difference (%)), and y-axis values represent the time lag (ms) that occurred when postural sway was synchronized. These data come from the Open-Open condition (see Fig 1 for the test conditions) and account for postural sway along an anterior-posterior axis only. As the figure shows, when the time lag is closer to 0, the ΣNCR difference (%) is closer to 0, i.e., the level of influence that paired participants exerted on one another was close to the same. This observation was supported by the regression analysis between the two factors (see the regression line included in the figure).
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4580648&req=5

pone.0137126.g005: Relation between participants’ influence and time lag for postural sway synchronization.Each dot in the figure represents the average of all trials for each of the paired participants. The data are plotted in the following way. X-axis values represent differences in paired participants’ influence on one another’s postural sway (estimated by ΣNCR difference (%)), and y-axis values represent the time lag (ms) that occurred when postural sway was synchronized. These data come from the Open-Open condition (see Fig 1 for the test conditions) and account for postural sway along an anterior-posterior axis only. As the figure shows, when the time lag is closer to 0, the ΣNCR difference (%) is closer to 0, i.e., the level of influence that paired participants exerted on one another was close to the same. This observation was supported by the regression analysis between the two factors (see the regression line included in the figure).

Mentions: A three-way repeated measures ANOVA with factors SENDER (Open, Blindfold), RECEIVER (Open, Blindfold) and DISTANCE (Near, Far) was carried out for LR and AP axes, separately (for the statistical results, see Table 1; for the results for an AP axis, see Fig 4). For both AP and LP axes, there was a significant interaction between SENDER and RECEIVER and between RECEIVER and DISTANCE (see Table 1(A)). Because of the significant interactions, two-way repeated measures ANOVAs with the factors DISTANCE and SENDER were conducted for Open and Blindfolded RECEIVER for both AP and LP axes. For the AP axis, effects of SENDER and DISTANCE were significant for Open RECEIVER (see Table 1(B)). The significant effect of SENDER found here implies that paired participants being OPEN, i.e., both receiver and sender being sighted, exert great influence on each other in postural sway synchronization. As for the significant effect of DISTANCE, such a result suggests greater importance for the receiver role (i.e., the receiver’s contribution to postural synchronization) in the Near DISTANCE condition than in the Far DISTANCE condition, for open, as opposed to blindfolded, sender. This pattern of the results was also supported by the results of a simple regression analysis between the difference in ΣNCR for paired participants and the time lag in their postural synchronization (see Fig 5). (Note: For the LR axis, the Blindfolded RECEIVER showed a significant interaction with SENDER and DISTANCE. This effect seems to be a random effect, considering all other results reported in this paper. However, it should be noted that further investigation is needed.)


Unintentional Interpersonal Synchronization Represented as a Reciprocal Visuo-Postural Feedback System: A Multivariate Autoregressive Modeling Approach.

Okazaki S, Hirotani M, Koike T, Bosch-Bayard J, Takahashi HK, Hashiguchi M, Sadato N - PLoS ONE (2015)

Relation between participants’ influence and time lag for postural sway synchronization.Each dot in the figure represents the average of all trials for each of the paired participants. The data are plotted in the following way. X-axis values represent differences in paired participants’ influence on one another’s postural sway (estimated by ΣNCR difference (%)), and y-axis values represent the time lag (ms) that occurred when postural sway was synchronized. These data come from the Open-Open condition (see Fig 1 for the test conditions) and account for postural sway along an anterior-posterior axis only. As the figure shows, when the time lag is closer to 0, the ΣNCR difference (%) is closer to 0, i.e., the level of influence that paired participants exerted on one another was close to the same. This observation was supported by the regression analysis between the two factors (see the regression line included in the figure).
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4580648&req=5

pone.0137126.g005: Relation between participants’ influence and time lag for postural sway synchronization.Each dot in the figure represents the average of all trials for each of the paired participants. The data are plotted in the following way. X-axis values represent differences in paired participants’ influence on one another’s postural sway (estimated by ΣNCR difference (%)), and y-axis values represent the time lag (ms) that occurred when postural sway was synchronized. These data come from the Open-Open condition (see Fig 1 for the test conditions) and account for postural sway along an anterior-posterior axis only. As the figure shows, when the time lag is closer to 0, the ΣNCR difference (%) is closer to 0, i.e., the level of influence that paired participants exerted on one another was close to the same. This observation was supported by the regression analysis between the two factors (see the regression line included in the figure).
Mentions: A three-way repeated measures ANOVA with factors SENDER (Open, Blindfold), RECEIVER (Open, Blindfold) and DISTANCE (Near, Far) was carried out for LR and AP axes, separately (for the statistical results, see Table 1; for the results for an AP axis, see Fig 4). For both AP and LP axes, there was a significant interaction between SENDER and RECEIVER and between RECEIVER and DISTANCE (see Table 1(A)). Because of the significant interactions, two-way repeated measures ANOVAs with the factors DISTANCE and SENDER were conducted for Open and Blindfolded RECEIVER for both AP and LP axes. For the AP axis, effects of SENDER and DISTANCE were significant for Open RECEIVER (see Table 1(B)). The significant effect of SENDER found here implies that paired participants being OPEN, i.e., both receiver and sender being sighted, exert great influence on each other in postural sway synchronization. As for the significant effect of DISTANCE, such a result suggests greater importance for the receiver role (i.e., the receiver’s contribution to postural synchronization) in the Near DISTANCE condition than in the Far DISTANCE condition, for open, as opposed to blindfolded, sender. This pattern of the results was also supported by the results of a simple regression analysis between the difference in ΣNCR for paired participants and the time lag in their postural synchronization (see Fig 5). (Note: For the LR axis, the Blindfolded RECEIVER showed a significant interaction with SENDER and DISTANCE. This effect seems to be a random effect, considering all other results reported in this paper. However, it should be noted that further investigation is needed.)

Bottom Line: As for the causality analysis, noise contribution ratio (NCR), the measure of influence using a multivariate autoregressive model, was also computed to identify the degree to which one's postural sway is explained by that of the other's and how visual information (sighted vs. blindfolded) interacts with paired participants' postural sway.It was found that for synchronization to take place, it is crucial that paired participants be sighted and exert equal influence on one another by simultaneously exchanging visual information.Furthermore, a simulation for the proposed system with a wider range of visual input showed a pattern of results similar to the behavioral results.

View Article: PubMed Central - PubMed

Affiliation: Division of Cerebral Integration, National Institute for Physiological Sciences, Okazaki, Aichi, Japan.

ABSTRACT
People's behaviors synchronize. It is difficult, however, to determine whether synchronized behaviors occur in a mutual direction--two individuals influencing one another--or in one direction--one individual leading the other, and what the underlying mechanism for synchronization is. To answer these questions, we hypothesized a non-leader-follower postural sway synchronization, caused by a reciprocal visuo-postural feedback system operating on pairs of individuals, and tested that hypothesis both experimentally and via simulation. In the behavioral experiment, 22 participant pairs stood face to face either 20 or 70 cm away from each other wearing glasses with or without vision blocking lenses. The existence and direction of visual information exchanged between pairs of participants were systematically manipulated. The time series data for the postural sway of these pairs were recorded and analyzed with cross correlation and causality. Results of cross correlation showed that postural sway of paired participants was synchronized, with a shorter time lag when participant pairs could see one another's head motion than when one of the participants was blindfolded. In addition, there was less of a time lag in the observed synchronization when the distance between participant pairs was smaller. As for the causality analysis, noise contribution ratio (NCR), the measure of influence using a multivariate autoregressive model, was also computed to identify the degree to which one's postural sway is explained by that of the other's and how visual information (sighted vs. blindfolded) interacts with paired participants' postural sway. It was found that for synchronization to take place, it is crucial that paired participants be sighted and exert equal influence on one another by simultaneously exchanging visual information. Furthermore, a simulation for the proposed system with a wider range of visual input showed a pattern of results similar to the behavioral results.

No MeSH data available.


Related in: MedlinePlus