Testing theories of reciprocity: do motivations matter?