Three fixed low-frequency signals do not improve late Kepler prediction for KIC 8164262 in eight fit variants

We test whether three published frequencies near 0.28 cycles per day give a stable light signal in KIC 8164262. We fit Kepler release 25 quarters 1 to 8 and predict quarters 9 to 17. We use 51,347 quality-zero samples outside an orbital mask. The early signal increases late mean squared error in all eight flux, trend, and mask fits. A quarter-quadratic fit gives a late root mean squared error of 186.03 ppm with the early signal and 185.77 ppm without it. A joint coefficient-change test gives conditional p values from 0.039 to 0.760 across fit and noise-group choices. The weak results depend on these choices. They do not establish either a stable signal or a change. Small fitted frequency shifts do not change this limit. The data do not distinguish Rossby modes, gravity modes, spots, or an instrument effect. The test uses one telescope. The published frequencies were selected from the same mission, so this is not independent confirmation. We report no new planet or stellar mode.

Kepler · KIC 8164262 · Rossby modes · coherence · null result