>>938907007
>Many studies suggest
>depending on the methodology
so if you play with the numbers enough it says what works for the argument you want to have?