> If two monkeys in one group died under anesthesia, that doesn't constitute ground to conclude they were more likely to die
I didn't say that two monkeys constituted statistical significance.
> Hence you may subtract them from the age related deaths.
You can subtract them because they're not statistically significant, but you can't conclude that there isn't something going on, which is what "there's no reason to suppose" means.
Suppose that odds of death due to anesthesia is really low (which it is) but that calorie restriction increased it by 10x (which we don't know either way). 10x a really small number is small enough that this study can't observe the effect because it doesn't kill enough monkeys. And yet, that 10x can easily cause two deaths in one pool while the other has 0.
I didn't say that two monkeys constituted statistical significance.
> Hence you may subtract them from the age related deaths.
You can subtract them because they're not statistically significant, but you can't conclude that there isn't something going on, which is what "there's no reason to suppose" means.
Suppose that odds of death due to anesthesia is really low (which it is) but that calorie restriction increased it by 10x (which we don't know either way). 10x a really small number is small enough that this study can't observe the effect because it doesn't kill enough monkeys. And yet, that 10x can easily cause two deaths in one pool while the other has 0.