A common stellar radius changes 10 of 1,031 listed planet size orders under linear scaling in the 21 October 2026 archive sample

We tested how different stellar radius scales affect the listed size order of transiting planets at the same star. We used two NASA Exoplanet Archive responses saved on 21 October 2026 UTC. Exact matches to planet radius source rows gave 1,031 consecutive transiting planet pairs at 692 hosts with one listed star. We held each source-implied planet-to-star radius ratio fixed and put both planet radii on a common stellar scale. This linear change reversed ten central-value size orders at ten hosts. All ten pairs used different planet radius sources. Seven changed from a larger outer planet to a smaller outer planet; three changed in the other direction. The larger-outer fraction changed from 726/1,031 (70.42%) to 722/1,031 (70.03%). Equal planet and star references within both source rows left 928 pairs and eight reversals. No reversal remained among 656 pairs whose two listed radius errors were at most 20%. A separate program checked the pair decisions with exact decimal cross-products. Common-source archive comparisons supported some flags but disagreed with the scaled order for Kepler-381. The result isolates a stellar scale effect in a fixed catalog sample. It does not establish true planet size order or a new planet. The package preserves the raw responses, source matches, exclusions, code, numerical results, figures, and offline repeat instructions.

exoplanets · planet size order · stellar radius · NASA Exoplanet Archive · data provenance · transiting planets · reproducible research