Beyond Synchrony: Joint Action in a Complex Production Task Reveals Beneficial Effects of Decreased Interpersonal Synchrony

Abstract

A variety of joint action studies show that people tend to fall into synchronous behavior with others participating in the same task, and that such synchronization is beneficial, leading to greater rapport, satisfaction, and performance. It has been noted that many of these task environments require simple interactions that involve little planning of action coordination toward a shared goal. The present study utilized a complex joint construction task in which dyads were instructed to build model cars while their hand movements and heart rates were measured. Participants built these models under varying conditions, delimiting how freely they could divide labor during a build session. While hand movement synchrony was sensitive to the different tasks and outcomes, the heart rate measure did not show any effects of interpersonal synchrony. Results for hand movements show that the more participants were constrained by a particular building strategy, the greater their behavioral synchrony. Within the different conditions, the degree of synchrony was predictive of subjective satisfaction and objective product outcomes. However, in contrast to many previous findings, synchrony was negatively associated with superior products, and, depending on the constraints on the interaction, positively or negatively correlated with higher subjective satisfaction. These results show that the task context critically shapes the role of synchronization during joint action, and that in more complex tasks, not synchronization of behavior, but rather complementary types of behavior may be associated with superior task outcomes.

Fig 1. Overview of the experimental setup and time-course.

(a) Setting in which participants built the cars and the location of the measurement devices. (b) Time-course of the experiment.

Fig 2. Illustration of MdRQA on artificial continuous signals, i.e., sine waves with added random noise.

Fig 2a depicts a set of three sine waves that exhibit a relatively low degree of synchrony together with their associated recurrence plot (2b). Fig 2c depicts a set of three sine waves that exhibit a relatively high degree of synchrony with their associated recurrence plot (2d). The recurrence plot in 2b is sparsely populated compared to 2d. Furthermore, the recurrence points are much less connected in 2b compared to 2d, as can be quantified by %Determinism. %Determinism is defined as the number of points in a recurrence plot that form adjacent diagonal line structures divided by the number of all points on the plot: while the plot in 2b exhibits %Determinism = 18.6%, the plot in 2d exhibits %Determinism = 54.1%. This figure was adapted from [7] with permission from Elsevier and Copyright Clearance Center (license number: 3926230267291).

Fig 3. Illustration of MdRQA on artificial stochastic signals, i.e., random numbers drawn from a uniform distribution.

Fig 3a depicts a set of three signals that exhibit a relatively low degree of synchrony together with their associated recurrence plot (2b). Fig 2c depicts a set of three signals that exhibit a relatively high degree of synchrony with their associated recurrence plot (2d). The recurrence plot in 2b is sparsely populated compared to 2d. Furthermore, the recurrence points are much less connected in 2b compared to 2d, which can again be quantified by the measure %Determinism, which is defined as the number of points in a recurrence plot that form adjacent diagonal line structures divided by the number of all points on the plot: while the plot in 2b exhibits %Determinism = 7.7%, the plot in 2d exhibits %Determinism = 33.6%.

Fig 4. Example time series and recurrence plots from one dyad’s hand movement accelerations (top panel) and heart rates (bottom panel) during each of the three conditions: EC (left panel), HC (middle panel), and TT (right panel).

(a) Hand movement accelerations and their recurrence plot during EC. (b) Hand movement accelerations and their recurrence plot during HC. (c) Hand movement accelerations and their recurrence plot during TT. The four time series displayed in (a-c) present the acceleration of participant A’s right hand, participant A’s left hand, participant B’s right hand, and participant B’s left hand, in that order from top to bottom. (d) Heart rates and their recurrence plot during EC. (e) Heart rates and their recurrence plot during HC. (f) Heart rates and their recurrence plot during TT. Please note that the recurrence plots might not present an accurate visualization, as the resolution with respect to the underlying data points is too coarse-grained.

Fig 5. Synchrony in hand movement accelerations as a function of building condition (EC, HC, TT) and data type (real pairs, false pairs).

Panel a) displays the results for %Determinism, panel b) displays the results for %Laminarity.

Fig 6. Effects of building condition (EC, HC, TT) on subjective perceptions for a) fun, b) difficulty, c) cooperation; d) power asymmetry, and e) product satisfaction.

Lines marked with * denotes p < .05, ** denotes p < .01, *** denotes p < .001, and n.s. denotes mean differences that are not significant (i.e., p > .05).

Fig 7. Effects of building condition (EC, HC, TT) on product outcomes for a) number of pieces used, and b) aesthetic appeal.

Lines marked with * denotes p < .05, ** denotes p < .01, *** denotes p < .001, and n.s. denotes mean differences that are not significant (i.e., p > .05).

Fig 8. Schematic summary of the effects of synchrony in hand movement acceleration and heart rate on subjective perceptions and objective product outcomes by building condition.