Investigating the effect of covert practice modality on improvements in motor performance

The goal of this study is to evaluate the differential contributions to motor skill acquisition of action observation (AO; watching someone else perform a movement), kinesthetic motor imagery (KMI; the mental rehearsal of a movement focusing on the sensations of the movement), and visual motor imagery (VMI; the mental rehearsal of a movement focusing on the picture of 'video replay' of the movement). The overarching aim is to advance knowledge concerning modalities of covert practice. Covert practice, including AO, KMI, and VMI, is thought to lead to the generation of an internal simulation of a movement and there is evidence that covert practice methods facilitate motor skill acquisition. Yet, the separation of these modalities and in turn knowledge of 'what' information contributes to motor skill acquisition and their relative effectiveness is not well understood. Stemming from evidence from behavioural and neuroimaging studies, the Motor Simulation and Performance Model proposes that different forms of covert practice vary in impact depending on skill level (McNeill et al., 2019). AO is predicted to be more effective than motor imagery (generally, with type of imagery not considered) for novices. While mechanisms and applications of AO and MI (typically KMI) have been compared in several studies (e.gs., Romano-Smith et al., 2018; Smith et al., 2019; Wright et al., 2022), no work has directly compared AO, KMI, and VMI within the same experiment, leaving a gap in understanding how these modalities of covert practice differ in their early effects on performance and of their overall separability. There is evidence that combining VMI and KMI yields additive benefits, suggesting that these modalities provide partially distinct contributions to performance (Callow et al., 2017). Studies investigating their application, typically test behavioural change following VMI or KMI separately. For instance, VMI has been shown to benefit tasks such as high jumping and ballet movements (Olsson et al., 2008b; Girón et al., 2012), whereas KMI is shown to facilitate motor skill acquisition in tennis and golf (Guillot et al., 2012; 2013; Smith et al. 2008). While these differences may reflect the task-specific nature of VMI and KMI (i.e., that VMI may be beneficial for tasks where technique, spatial trajectories, and dynamic adjustments are important, whereas KMI may benefit tasks that focus on the core execution of movements where motor control, and precision is important), due to the variability in efficacy of MI types across tasks, the functional separation between VMI and KMI remains unclear. Interestingly, work by Peters et al. (2025) examined MI performance improvements on a novel gesture sequence task by manipulating experience (physical or observational practice groups) and assessing these groups’ independent VMI and KMI performances. In showing better performances for the observational practice group for both VMI and KMI instructions, the authors postulated a possible reliance of KMI on a visual representation formed through VMI (i.e., that ‘execution’ components of KMI require an initial visual representation; Peters et al., 2025). Without this visual representation, novice participants may need to rely on a visual (motor) image (Olsson et al., 2008a) until adequate ‘scaffolding’ for KMI has been created. While speculative, it thus follows that initial acquisition of a novel skill may be more effective when practiced through VMI than KMI. Accordingly, the current study was conducted to test the differential contributions to motor skill acquisition of action observation (AO; watching someone else perform a movement), kinesthetic motor imagery (KMI; the mental rehearsal of a movement focusing on the sensations of the movement), and visual motor imagery (VMI; the mental rehearsal of a movement focusing on the picture of 'video replay' of the movement); first, to test predictions of the Motor Simulation and Performance Model that AO will lead to greater skill acquisition than either form of motor imagery of a novel task (i.e., in novices), and second, to further test the separation between KMI and VMI. In particular, if KMI relies on an initial visual representation formed through VMI, then VMI should lead to greater skill acquisition than KMI of a novel task.