Comparison-induced distortion theory (Choplin 2007; Choplin and Hummel 2002) describes how comparison words like “better” suggest quantitative differences between compared values. When a comparison word is used to contrast a personal attribute value with some standard (e.g. “Your score is better than average”), the comparison-suggested difference for the word may bias estimates or recall of personal attribute values. Three studies investigated how comparison-suggested differences determine the effect of social comparison on estimates or recall of personal attribute values. The first study demonstrated that estimates of attributes are biased towards (assimilation) or away from (contrast) a comparison standard depending on whether the difference between the compared attribute values exceeds or falls below the comparison-suggested difference. The second study showed that the comparison language selected by participants (through the difference suggested by the language) mediated the effect of standard similarity on attribute estimates following a social comparison. The third study demonstrated concurrent assimilation and contrast effects in recall of attribute values due to the size of the observed difference between the self and the standard for the attribute. Unlike in previous research on social comparison, assimilation and contrast patterns in these studies can be explained through a single process.