Gravar-mail: Golden Rule or valence matching? Methodological problems in Hamlin et al.