The extent to which different raters or observers produce the same results when observing and coding the same data.