Good indicators should be redundant measures of the same thing, not complete

Evidence

Connecting the data–indicator level and the secondary level in figure 6.1 is the “indicator of” relationship. In general, the Completeness and Redundancy Guidelines do not apply here. In fact, it is often the opposite: indicators are ideally redundant, i.e., measures of the same thing. One does not want a “complete” set of indicators, but a reasonable number of good (e.g., highly correlated) indicators. — Gary Goertz, Social Science Concepts and Measurement, 928

Categorization

• DEFINE: Clarifies the ‘indicator of’ relationship and inverts the usual completeness intuition for measurement.
• XREF: Connects to psychometric reliability logic (convergent validity) and to the book’s own Completeness/Redundancy Guidelines. — connects to: Redundant status signals risk looking too intentional and boring, so overdoing consistency backfires. — W. David Marx, Status and Culture: How Our Desire for…, 230