Only one of five absolute sex-gap ozone contrasts kept its sign under a 1% mean-error bound in a 2024 extract

We tested whether ozone changed the absolute gap between female and male development times in one direction across a 2024 invertebrate extract. Ten sex-labelled rows gave five matched comparisons from three papers. Gaps narrowed in three comparisons and widened in two. A Canadian 1992 comparison changed sex order while its absolute gap narrowed, so a signed interaction alone gives an incomplete description. Only one absolute-gap direction remained fixed when each stored mean could change independently by 1%. We give exact attainable bounds and verify them with linear programs. The 1% allowance is assumed, not measured. Complete sex reporting, survival selection, missing covariance and shared US facilities limit inference. The extract has no independent region-by-period replication grid. No climate trend, temperature interaction, adult mating effect or statistical significance is established. Code, aggregate results, source hashes and external-data retrieval instructions are included.

ozone · ecological timing · sex differences · observation bias · development time · sensitivity analysis · absolute gap