Dichotomous poverty indicators function as weighted interval-level factors.

Evidence

The Multidimensional Poverty Index illustrates how this works in a multilevel and multidimensional setting. At the core, bottom level of the concept are various data about dimensions such as health, education, or living standards. These basic data are dichotomous and aggregated usually via the maximum to produce 0 or 1 values for each of the 10 core indicators. Are these dichotomous indicators nominal, ordinal, or interval? The basic-level numeric value—the poverty measure—is the weighted sum of these 10 indicators. So the 0 or 1 for each indicator is an interval-level factor which is then weighted to produce the final sum. So what might appear to be a nominal indicator is an ordered [0, 1] variable which when weighted becomes an interval-level value for the degree of poverty of a given individual or household. — Gary Goertz, Social Science Concepts and Measurement, 775

Categorization

• XREF: Connects to broader work on composite indices and the Alkire-Foster method for multidimensional poverty measurement.
• SHARE: A neat technical curio for anyone working with index construction — that 0/1 indicators secretly function as interval factors once weighted.