Nucleation-assisted fluid inclusion microthermometry (NA-FIM) provides direct estimates of speleothem formation temperatures (<em>T<sub>form</sub></em>), derived statistically from measured homogenization temperatures (<em>T<sub>h</sub></em>). Accurate analysis requires inclusions to act as closed, isochoric systems. However, inclusions frequently undergo post-entrapment alteration via three mechanisms: mechanical stress during preparation, thermal re-equilibration, and direct laser impact during analysis.</p> <p>While severe alteration produces visually obvious leakage, cryptic alteration irreversibly expands inclusion volumes without visual evidence. This expansion systematically lowers fluid density, causing initially symmetrical <em>T<sub>h</sub></em> distributions to distort into asymmetric, right-skewed ones. Consequently, traditional statistical metrics (the mean, median, standard deviation, and standard error) become mathematically untenable and systematically overestimate formation temperatures.</p> <p>To robustly estimate <em>T<sub>form</sub></em> from positively biased data, we propose uniformly applying the first quartile (Q1) of the measured <em>T<sub>h</sub></em> distribution. Numerical simulations demonstrate that Q1 closely approximates the true pristine mean under variable right-skew conditions. To accurately capture uncertainty without relying on parametric assumptions, we recommend reporting overall data scatter using the interquartile range (IQR) and quantifying the precision of the Q1 estimate via bootstrap 95 % confidence intervals. This robust approach safeguards the integrity of paleotemperature reconstructions.