Typically, laboratory experiments suffer from homogeneous subject pools and self-selectionbiases. The usefulness of survey data is limited by measurement error and by thequestionability of their behavioral relevance. Here we present a method integrating interactiveexperiments and representative...