factual

Measuring an unobservable construct with multiple methods

Because tightness-looseness is not directly observable, validating it requires multiple measures—surveys of the permissible range of behavior across situations and countries, coding ethnographies of preindustrial societies, dictionaries, brain imaging, and computational models—seeing whether predictions are borne out homologously across levels.

factualpending

Speaker

Michele Gelfand

Evidence Quote

“you need to provide validity that this construct is measuring what it purports to measure because it's not observable necessarily. And my approach... is to use multiple measures.”

Source

Michele Gelfand on Rule Makers, Rule Breakers— EconTalk
Created: 6/13/2026, 12:12:12 AM

Your notes

Loading notes...