Erratum: Studying Identities with Experiments: Weighing the Risk of Posttreatment Bias Against Priming Effects (Journal of Experimental Political Science (2019) DOI: 10.1017/XPS.2019.26)