Field guide for Supplementary Material 2

subject_id / subject: released CoSpine participant identifier.
run: motorL or motorR task run; not HCP phase-encoding direction.
roi: M1, S1, PMd or PMv, using the fixed Brainnetome definitions in Methods 2.4.
hemisphere: anatomical left/right hemisphere.
include: saved input/coverage qualification decision; see reason for exclusions.
input_PNM_dependency: saved dependency assessment for the run, not a new PNM model.
n_volumes / TR: retained BOLD volume count / repetition time in seconds.
valid_voxels / atlas_voxels: effective ROI voxel count / full parcel voxel count.
coverage_percent: 100 times valid_voxels divided by atlas_voxels; threshold 80%.
design_rank / dof: saved first-level design rank / residual degrees of freedom.
D_M1 / D_S1 / D_PMd / D_PMv: hand-aligned regional COPE differences.
D_motor: equally weighted four-region average, calculated within participant.
right_hand_contra_minus_ipsi / left_hand_contra_minus_ipsi: hand-specific differences.
endpoint: original machine-readable identifier for a regional or composite measure.
n / effective_n: actual participant count for the specified measure/comparison.
mean / mean_paired_difference: saved group estimate in the cohort's COPE units.
sd / sd_paired_difference: participant-level sample standard deviation.
cohens_dz: mean divided by the sample standard deviation.
bootstrap_ci_low / bootstrap_ci_high: unadjusted percentile 95% interval limits.
positive_n / negative_n / zero_n: counts for the specified participant-level measure.
exact_p_two_sided / raw_p: unadjusted p value for the stated test_method/exact_test.
holm_adjusted_p: adjusted p value within the stated family, not across all files.
reject_holm_0.05 / reject_at_0.05: saved decision under the specified family/test.
weights_M1_S1_PMd_PMv: retained within-participant regional contrast weights.
dataset / result_set / analysis_role: source cohort, analysis set or historical role;
  use the README and Methods rather than a historical role label alone.
WTS: covariance-aware regional equality-of-means Wald-type statistic.
p / new_p in WTS files: plus-one parametric-bootstrap p value.
bootstrap_resamples / permutation_count / replicates: number used by that procedure.
seed / bootstrap_seed: saved random-number seed for that procedure.
extreme_count: count at least as extreme as observed, under the stated comparison rule.
coefficient_L_minus_R: observed-order effect, left-hand-first minus right-hand-first.
HC3_SE / ci_low / ci_high: HC3 standard error / interval limits for the order coefficient.
test_statistic_F / p_value in the run-order file: regional-profile-by-order statistic
  and its participant-label permutation p value, not a between-cohort test.
A/B/C/FINAL fields: see the explicit sample/processing decomposition in the README.

HCP_five_contrasts.csv retains a broad historical schema. Blank omnibus or mixed-model
fields do not supply additional inference; its five populated regional-comparison
rows are the evidence used here. Participant-level HCP data and identifiers are not
included, so this archive alone cannot reproduce HCP participant-level figures.
