Everyone assumes things that aren't true, not just "junior" data scientists. There is not standard methodology for non gaussian non independent random variables.