Measuring Norms: Eliciting normative expectations with Krupka and Weber's method allowing for neutral reports

Abstract

This study examines whether adding a neutral category improves Krupka-Weber’s (2013) coordination game approach for eliciting normative expectations. The standard four-item scale lacks a neutral midpoint, potentially obscuring asymmetries in deviations from prescribed actions and forcing expression of normative valence, even when absent. In experiments using dictator games and vignettes, we introduce a neutral option. This addition neither serves as a source of focality -originally a concern for coordination-based methods nor distorts responses, and remains robust when combined with incentivized uncertainty response options. Including it slightly reduces bias and noise in predicting behavior, enhancing precision and interpretability without introducing drawbacks. 

Keywords: social norms, normative expectations, focal points, coordination, neutral, experimental methods

JEL Classification: A13, C91, D01, D63, D83, D91