Pointing terms reduce median source-response differences for two weak KIC 5520878 signals in Kepler Q2 to Q16

We test image responses of two known weak signals of KIC 5520878 in Kepler release 25 quarters 2 to 16. With 20-day local nuisance fits, adding reported pointing terms reduces the median response difference from the primary pulsation from 0.078 to 0.041 pixel for the 1.12535 cycles/day signal and from 0.095 to 0.038 pixel for the 1.13577 cycles/day signal. Fits with 10- and 40-day windows retain a lower median with pointing terms, but the response values change. Conditional errors and a linear transfer check are supplied. The result tests source discrimination under a fixed image model. It does not identify a separate source, validate a stellar mode, or establish a new planet. Flux and image centers share Kepler images. There is no independent instrument confirmation. Code, selected samples, source hashes, full tables, and reproduction instructions are included.

Kepler · KIC 5520878 · RR Lyrae · image centroid · source contamination