## Introduction

### The serial reaction time task

### Metre

### The current study

## Method

### Participants

### Materials and procedure

#### Serial reaction time task

`lmer`

from the `lme4`

package (Bates et al., 2014) in R], after ensuring model residuals were normal via diagnostic procedures (`DHARMA`

package Hartig (2019)). Although it has been recommended in the psychology literature to use generalised linear mixed effects models for reaction time data in general (Lo & Andrews, 2015), this advice should in fact only apply when the distribution of model residuals, and not of the raw data, is non-normal (Kéry & Hatfield, 2003). Model term selection was guided by Akaike information criterion [AIC;Sakamoto et al. (1986)]. t-Statistics with Satterthwaite degrees of freedom calculated the significance of fixed effects (`car`

package (Fox et al., 2012)). For each analysis, full details concerning the model and its specification, including random intercepts and slopes, are reported in the Supplementary Materials. Significant interactions were further investigated using Tukey’s honestly significant difference (HSD) tests of estimated marginal means [`emmeans`

, Lenth et al. (2018) in R]. The semi-partial \(R^2\) (R\(^2_{sp}\)) is reported as an indication of relative magnitude of effect sizes (Jaeger et al., 2017). Confidence intervals for the regression coefficients and R\(^2_{sp}\) were both calculated in R using `Confint`

and `R2beta`

(Jaeger et al., 2017), respectively. Other pairwise contrasts were conducted using t-tests (`ttest2`

in MATLAB) or Mann-Whitney U-tests (`ranksum`

in MATLAB), and linear trends were inspected with Pearson’s linear correlation coefficient or Spearman’s rank coefficient (both `corr`

in MATLAB). The threshold for statistical significance was 0.05 and multiple comparison corrections were applied to post hoc contrasts using the Bonferroni method. The false discovery rate (FDR; Benjamini-Hochberg procedure) was used to correct correlations and the permutation analyses in the exploratory section. In all cases, adjusted p-values are reported. Descriptive statistics (e.g., Mean, SD) reported at the group level are first aggregated within Participant.## Results

### Summary statistics

_{3/4 m}\(= 3\), N

_{4/4 m}\(= 2\)) produced more than 10% correct anticipatory responses across multiple blocks and were excluded from further analysis, resulting in a final n of 41 (3/4 m condition \(= 17, 41.5\%\), 4/4 m condition \(= 24, 58.5\%\)). The remaining correct anticipatory responses (mean \(= 2.24\%\), SD \(= 1.84\%\)) produced by other participants were removed. Overall, there were 49,785 (93%) correct responses retained. Mean percent correct did not vary by Metre condition, U(N

_{3/4 m}\(= 17\), N

_{4/4 m}\(= 24\)) = 347.5, z = – 0.24, p \(= 0.81\)) and did not correlate with Rhythm Score (r

_{s}\(= 0.01\), p \(= 0.94\)). Calculated within participants (41), the mean of mean RT was 430.98 ms, 95% CI [401.23,460.74], SD = 94.27, Range = [247.01,833.66]. A visualisation of the complete task is shown in Fig. 4a. Descriptive statistics concerning self-reported musical background and performance in the Rhythm Discrimination Task are given in Table 1, the measures of which were correlated, r\(_\mathrm{s} = 0.65\), CI [0.43, 0.79], p \(< 0.001\). Median levels of instrument playing did not statistically differ between the 3/4 m (median \(=\) 2.5, IQR \(=\) 2.5) and 4/4 m (median \(=\) 2, IQR \(=\) 2) conditions, U(N

_{3/4 m}\(= 17\), N

_{4/4 m}\(= 24\)) = 310, z = -0.05, p \(= 0.96\). Rhythm Score, defined as percent correct in the Rhythm Discrimination Task, ranged from 42.5 to 100, with a median score of 85. As regards the distribution of Rhythm Score across Metre conditions, the 3/4 condition (median \(=\) 87.5, IQR \(=\) 15.38) did not significantly differ in comparison with the 4/4 condition (median \(=\) 83.75, IQR \(=\) 22.5), U(N

_{3/4 m}\(= 17\), N

_{4/4 m}\(= 24\)) = 390.5, z = 0.88, p = 0.38.

Self-reported musical experience | ||||
---|---|---|---|---|

Rating | Count | Percent | Rhythm Score | |

Mean | SD | |||

1 | 4 | 10.5 | 66.88 | 18.19 |

2 | 15 | 39.48 | 77.5 | 12.85 |

3 | 7 | 18.42 | 79.29 | 18.47 |

4 | 8 | 21.05 | 87.5 | 6.55 |

5 | 4 | 10.5 | 92.5 | 7.91 |