Rating versus ranking: What is the best way to reduce response and language bias in cross-national research?