For fixed auditory tables, exact sampling risks favored a common estimate over means from four sounds per class in 19 of 21 iEEG patients and 0 of 24 fMRI adults

We tested whether acoustic correction improves short maps of auditory class means. Separate public studies supplied human iEEG voltage and BOLD fMRI. Four patients and four MRI adults supplied development. A local code lock preceded new endpoint scores in 21 patients and 16 other MRI adults. Each panel used four sounds per class across five classes. A split acoustic rule corrected each half using slopes fit on the other half. Mean error gains were 0.224% in iEEG and 0.568% in primary fMRI. Their 95% person Student limits were [0.106,0.341]% and [0.526,0.610]%. Mean electrical skill versus one common calibration estimate was -61.586%, with limits [-89.192,-33.981]%. MRI skill was +42.396% [39.775,45.017]%. The shared gain-and-skill rule failed, p=0.999980. An added exact finite-library calculation compared ordinary class means with the common estimate. At four sounds per class, ordinary means had greater expected error in 19 of 21 electrical patients and none of 24 MRI adults from two studies. A closed form rule gave every feasible count threshold and oracle shrinkage factor. The added calculation is exploratory and conditional on the fixed processed tables. It does not identify a causal modality difference. Baseline subtraction left electrical gain uncertain. A one-site MRI control had negative skill. Full sound deletions, every negative result, direct regression checks and exact small-table enumeration remain. The package provides a checked estimation tool and negative cross-modal accuracy benchmark. It does not confirm a positive biological generalization, absent class coding, or prospective time savings. Code, permitted inputs, source pins and terms, full results, figures and repeat steps are supplied. Original MRI arrays stay external.

human iEEG · fMRI · auditory cortex · category means · finite population sampling · control variates · negative result