A coding supervisor monitors quality by having coders recode records done by peers. This practice is known as:

The practice of having coders recode records completed by their peers is referred to as interrater reliability. This term describes the degree of agreement or consistency between two or more raters or individuals assessing the same data or outcomes. In the context of coding, when one coder's work is reviewed and coded by another coder, interrater reliability is being evaluated. By doing this, the supervisor is ensuring that different coders arrive at similar conclusions for the same records, which is crucial for maintaining the integrity and accuracy of coding practices within healthcare settings.

Intrarater reliability would refer to the consistency of a single coder over time in coding the same records, which does not apply in this scenario where the focus is on peer evaluations. Similarly, interrater validity and intrarater validity are not applicable because they concern the accuracy of the measuring tool and whether it truly measures what it intends to measure, rather than the agreement between different raters. Thus, interrater reliability is the correct term that encapsulates the quality monitoring method described in the question.