They need to be careful they don't word their questions in a way that purposefully manipulates the outcome. That is already a problem and this research gives them a clearer way to do just that.