<p>Anonymized dataset underlying the study findings.</p>

Mixed-design ANOVAs examined pre-to-post trajectories.</p> <p>Results</p><p>FR-A had lower adjusted post-test reaction time than FR-B under best-trial (difference = −21.86 ms, 95% CI: −30.31 to −13.41; adjusted p < .01), mean-based (−12.36 ms, adjusted p < .01), and median-based scoring (−10.94 ms, adjusted p = .03). The FR-A–control difference was significant in the standard best-trial ANCOVA (−15.44 ms, 95% CI: −24.12 to −6.76; adjusted p < .01) and extended bootstrap comparison (−14.79 ms, adjusted p < .01), but not in mean-based (adjusted p = .33) or median-based analyses (adjusted p = .39).

Extended bootstrap estimates were also higher in FR-B than in the control group (9.52 ms, adjusted p = .01). Because this model contained baseline interactions, these contrasts represented average rather than constant differences across baseline values. The time × group interaction was significant only for best-trial scoring (p < .01).</p> <p>Conclusion</p><p>FR-A had consistently lower adjusted reaction time than FR-B across all scoring methods.

However, the FR-A–control comparison depended on baseline reaction time and scoring method. Thus, the findings do not establish a scoring-independent benefit of the 30-second protocol over passive rest or support a firm practical recommendation.</p></div>