We propose an integrative theoretical framework of the cognitive-ecological processes of impression updating. The framework combines Bayesian belief updating with the classical conceptualisation of mental scaling – the idea that cognitive representations of ecologically scaled values are systematically disproportional. A straightforward way to approximate the mapping of ecological scales (e.g., prices in €) on mental scales is to use ecologically determined relative ranks (e.g., Stewart et al, 2006). Four advice-taking (“judge-advisor-system”) experiments (N=672) replicated and consolidated effects of advice convergence and distance, as well as the amount of advice, while all advice properties were manipulated based on ecological frequency. We thereby replicated existent findings on advice properties while constantly and experimentally controlling for ecological frequency. In a second step, we consider that ecological frequencies are typically strongly skewed (mostly lower numbers are dense and frequent, while higher numbers are typically distinct and rare). We show that such ecological asymmetries impact how confidence ratings are made, as well as changes in estimate and confidence – even when we statistically control for distance on the ecological scale. Along these findings, we demonstrate that the inclusion of mental scaling serves well in generating novel predictions and overcoming overly narrow and often times implicit boundaries and auxiliary assumptions of (normative) models applied to impression updating tasks.