When Group Rankings Conceal Listener Disagreement

Auditory recordings allow landscape experiences to enter design discussion, but stable group averages may conceal changes in individual comparisons. Thirty listeners assessed four settings in stereo and binaural presentations using three seven-category presence items. This investigation calculated matched format differences, classified within-listener setting contrasts as retained, reversed or tied, and examined sensitivity to incomplete answers, scoring and presentation position. Complete responses supported 110 format pairs and 150 cross-setting comparisons. Average binaural-minus-stereo differences ranged from −0.47 to 0.17 points, with all listener-bootstrap intervals crossing zero. The ordering of setting means was unchanged, yet 41 comparisons reversed, corresponding to 27.3% of available comparisons and 32.5% of comparisons with strict orderings in both formats. Sixteen reversals retained contrasts of at least one point in both presentations. Equal listener weighting yielded 25.0%, while restriction to fully observed listeners yielded 30.8%. Median scoring reduced reversals and increased ties, demonstrating dependence on comparative resolution. The observations establish that preserved aggregate ordering does not guarantee preserved individual judgments. They do not identify a causal format effect because each setting was presented only once per format, and no on-site counterpart was assessed.