Reply To: P. 286 Prostatectomies

January 17, 2018 at 9:52 pm #19714
Mike Kim

Hi Sean,

Thanks so much for the comments and great to have you here!

You certainly aren’t the first to ask about that drill – it’s a doozy —

In terms of answer choice (C) – it basically tells us something about the standard deviation of traditional vs robotic surgeries —

I certainly understand your thinking, but there is just as much likelihood that traditional surgery was dragged down by a bunch of super-fast outliers, and, more importantly, we can’t make any necessary inferences about pace from this information, and that’s what we need to bridge this gap in the argument.

To explain with an analogy — imagine that on average it takes you 10 minutes to drive to work, and on average it takes me 20 minutes to drive to work.

We cannot conclude from this premise who drives faster on average — it could be that I have a much longer distance to travel than you do (or not) —

And that’s the same problem with the original prostatectomies argument — we can’t conclude that lower average time is analogous to being faster because we don’t know if the average times were impacted by the fact that they were for different types of prostatectomies that inherently take more or less time.

Going back to the driving analogy — imagine we got some additional information (the equivalent of answer c) — my drive times are much more inconsistent than yours. So, it could actually be that most days I get to work faster than you (say, in 8 minutes) but there are certain days when I get caught in traffic and it takes me an hour to get to work, driving my average up over yours —

This might have an impact on whether we can use the average speed premises to conclude whether, on most days, I get to work in less time than you do —

But it doesn’t tell us anything more about whether I drive, on average, at a faster or slower pace than you do — whether my times are inconsistent or not, this doesn’t impact my average pace — my average pace is still what it is, your average pace is still what it is, and, without any information about distances, we cannot conclude anytime about our average paces relate to one another (who drives at a faster speed, on average) just based on information about time.

In the same way, the inconsistency in time that (C) hints at doesn’t directly impact the average pace of robotic vs traditional surgery.

Man, that seemed so much simpler in my head than it does written out, but I hope that helps clear things up a bit, and, if it doesn’t, happy to discuss further —