Selected odor response reversals gave positive mean test scores in 315 traced sister cells from two mice

We asked if cells with a traced common glomerulus gave stable opposite orders for pairs of odors. We made a score with separate selection and scoring repeats. Two repeats selected opposite target and sister contrasts. Two other repeats supplied products between a target contrast and a sister contrast from different repeats. Under stable means and independent zero-mean errors between complete repeats, increasing functions of one common scalar give a nonnegative expected score. The score allows arbitrary within-repeat error covariance. We tested 315 cells in 52 groups from two anesthetized mice, with 47 odors and four repeats. Group-weighted mean scores were +0.1439 and +0.1460. All six repeat partitions gave positive mouse means. The equal-mouse mean was +0.1449; a conditional group resample range was +0.0825 to +0.2274. This range is not animal-population uncertainty. Planned response checks kept positive means. A pre-odor control gave a large negative score in one mouse and exposed baseline sensitivity. This is a directional diagnostic, not a calibrated significance test. A positive score does not establish a common scalar model. In a synthetic sparse alternative, 7.22 percent of true odor-pair orders were reversed, but the mean score remained positive. The contribution is the expectation proof, the measured-data test, and a reusable package. It extends earlier fixed linear gain tests. It does not establish a neural mechanism or exclude local order differences. The package preserves code, compact data, source checks, all scores, figures and repeat instructions.

olfactory bulb · structure and function · sister cells · common input · odor ordering · independent repeats · calcium imaging · measurement covariance