Don’t show me a junk study of less than 70 people and doggedly latch onto as ‘proof’ and then disregard a study of over million. A small study can be an important indication of further research needed. But the statistical relevance is apple and oranges