When to trust (and not to trust) peer reviewed science