, 2004), and in patients with psychogenic tremor (Edwards et al., 2011), suggesting they provide a valuable indication of the experience of volition. The standard deviation of repeated judgements provides an additional, independent
measure of experience, akin to phenomenal clarity and precision. For example, vague and variable phenomenology of volition should produce a high standard deviation of intention judgements, while a clear experience that reliably precedes actions by a fixed latency should produce a lower standard deviation. As a control for non-specific aspects of the Libet task, including using the rotating clock hand as a chronometric device for timing subjective see more experiences, Trichostatin A concentration we asked participants to perform an additional block of trials in which they judged the time of their actual keypress, rather than the
intention that caused it. Trial order was counterbalanced between the two judgement conditions. The means and standard deviations of 40 intention judgements and of 40 action judgements were estimated for each subject. To investigate the relation between tic behaviour and experience of volition, we used a multiple regression model to predict the mean time of intention across participants. We used a range of predictor variables covering two main domains: First, we included three tic-related predictors: overall actual tic severity (RF), premonitory urges (PUTS scores), and capacity for intentional suppression of tics (IP). In addition, we included two general,
non-tic-related factors likely to check details influence conscious intention. These were the degree of attention deficit (FBB-ADHS), and the reliability of each individual’s W judgement (SD W), which partly reflects the criterion used to judge the onset of intention. The detailed justification for each of these predictors is given in Supplementary Text 1. Finally, in order to assess, whether GTS has a specific effect on perception of intentions, without generally altering time estimation or perceptual judgement about other motor events such as actions, a separate regression was performed for judgements of the keypress action (M-judgement), using the same regression model as for judgements of intention. The experience of intention (mean of Libet’s W judgement) occurred at a similar time in patients (mean – 184 msec ± 147 SD) and controls (mean – 185 msec ± 97 SD). Also, the estimated time of the keypress (M-judgement) was comparable between patients (mean – 56 msec ± 56) and controls (mean – 68 ± 46). Comparison of volition measures between GTS patients and healthy volunteers yielded no significant effect of group (F1,55 = .094, NS). There was an expected difference between the perceived time of intention and the perceived time of action (F1,55 = 72.536, p < .001), but no interaction with group (F1,55 = .124, NS).