A fixed-profile test finds no row-time trend in 13 Kepler quarter 7 stars near two comet candidates

We test whether dip times vary with detector row near two reported comet candidates in Kepler quarter 7. We fit a fixed loss profile to 37 public light curves. The two candidates are excluded from the row-time test. In 13 other channel-79 stars, the main correlation is -0.118 and the label permutation probability is 0.6925. None of eight tested fit choices gives a probability below 0.05. A row model predicts reference times worse than a constant model. In a 2000-ppm injection check, nine of 13 centers are recovered within 0.1 day, but three errors exceed 0.2 day. The data do not confirm a constant-speed row disturbance under this profile and window. Weak timing information and incomplete injection recovery prevent a strong exclusion of additive detector effects. We report no new comet or planet.

Kepler · exocomets · instrument artifacts · rolling bands · null result