Lack of Replication or Generalization? Cultural Values Explain a Question Wording Effect

Abstract In the context of current “replication crisis” across sciences, failures to reproduce a finding are often viewed as discrediting it. This paper shows how such conclusion can be incorrect. 1981, Schuman and Presser showed that including word “freedom” in survey question significantly increased approval allowing speech against religion USA. New experiments probability sample surveys (n = 23,370) USA 10 other countries wording effect replicated appeared four (Canada, Germany, Taiwan, Netherlands) but not remaining countries. The only which value freedom is especially salient endorsed. Thus, public support for proposition was enhanced by portraying it embodying principle nation’s culture. Instead questioning initial findings, inconsistent results signal limits on generalizability identify an important moderator.