Positive person scores did not pass a shared typicality test for 173 objects in eight EEG adults and three fMRI adults

We tested whether the visual pattern relations of the same object follow its typicality ranks in different categories. Separate public studies supplied scalp EEG and BOLD fMRI. Two EEG adults supplied development. A fixed test used eight other EEG adults and three fMRI adults. It had 173 target objects and 383 category memberships. Each score removed object and category terms and adjusted for an early or lower visual measure. Mean partial correlations were 0.2869 and 0.3584. The 95% person t intervals were [0.2584,0.3149] and [0.2139,0.4876]. Every person score was positive. The shared person p value was 0.004716. The full shared rule failed. Joint resampling of people, panel objects and categories gave ranges [-0.1470,0.5294] and [-0.1255,0.6444]. Each modality had 23 undefined draws out of 1,999, above the fixed 1% limit. Those draws remain. Every single category or target-object omission kept positive means. That did not ensure a positive joint range. An exact constructed example gave r=0.866 with one common latent pattern and different observation reliability. It was not fitted to people. Thus positive person scores do not confirm the declared cross-modal generalization or a prototype mechanism. The person test assumes a suitable person distribution; three fMRI adults give weak support for that assumption. Separate cohorts, rapid EEG sequences, selected voxels, source processing, reference coverage and fixed external norms limit transfer. The contribution is a new assay and matched benchmark. The package preserves code, licensed distances, exact IDs, source versions and terms, all results, model arrays, figures, local plans and repeat instructions.

human neuroscience · EEG · fMRI · object recognition · typicality · category structure · reference uncertainty · measurement reliability