The synchronisation of interpersonal behaviours in everyday life is essential to achieve joint actions tasks. Even in social interactions where there is no specific coordination goal, particular spatio-temporal relations are maintained between individuals unintentionally and are an important factor in social bonding. Two distinct approaches have been proposed to understand coordination in interpersonal movements during social interactions: the embodied simulation approach links behavioural matching to shared neural resources, which are activated both when an action is observed and when it is performed. Recent studies suggest that only biological stimuli evoke action imitation, as non-biological stimuli are processed elsewhere in the brain. An alternative approach, the coordination dynamics perspective, does not specify any neural substrate but views synchronisation as an emergent phenomenon of an underlying dynamical process, in which the components of a system self-organise toward a stable state. From this standpoint, the dynamic patterns of coordination are affected by attributes of the stimulus kinematics irrespective of whether this stimulus is biological or non-biological in nature. In this thesis, I investigate whether motion kinematics that are perceived as human will facilitate interpersonal coordination more readily than stimuli perceived as artificial. First, in two psychophysical experiments participants were asked to distinguish between real human movements and artificially produced movements. The findings provide insights into the features of one-dimensional cyclic movements that allow them to be identified as human; specifically, observers perceived movements of a particular range of smoothness and frequency as human, whereas both, very fast or very slow movements outside this range were reliably distinguished as artificial. Second, these distinct subsets of human and artificial movement kinematics were applied as stimuli in subsequent intentional and unintentional coordination experiments, and variations in both the strength and pattern of coordination response was observed. In contrast to expectations derived from the embodied simulation literature, experiments did not provide any evidence that the perceptual identification of a stimulus as human matters with regard to improving coordination. Instead, all modulations in coordination behaviour could be explained solely on the basis of direct effects of different stimulus kinematics on an underlying dynamical system. A subsequent modelling study showed that the same patterns of coordination occur with a system of coupled oscillators when the same stimuli are applied. The consistency between the theoretical model and empirical results suggests that the observed coordination behaviour in human subjects can be explained on the basis of an underlying dynamical system, in accordance with the coordination dynamics approach, without the need to incorporate perceptual factors or specialised neural networks. Future studies will have to clarify whether factors relating to the perception of stimuli, predicted by embodied simulation, might become more important in the absence of larger scale effects associated with stimulus kinematics or with a more ecological stimulus.