You can lead a horse to water but…

March 2, 2021

Scientists are complete idiots on the business of science.

Once upon a time I asked Science Twitter to opine on whether they had ever met anyone smarter than they are. They, being mostly not sociopathic narcissists, said they had.

So far, so good. They are also quite willing to admit the obvious,

except for those weenie few per cent that are left after we discount the spoiler/joker rate. These are undoubtedly the duplicitous folks that claim never to think anyone else is an idiot.

I threw in a distractor for the parents in the crowd….

..but seriously, for many in my audience, this is going to be the situation where they are most able to rub elbows with something like the full distribution in their country. Public elementary school. Of course, many will already be on a very select track due to their choice of geographic location.

Next, we moved on to some science logic. Totes different and unrelated issue.

Here we see the joker rate of 3.4% on full display. Most Science Twitter types know this is nonsense. We LIVE for trying to assess the central tendency within a sample that expresses some portion of the variability that exists in the presumed population we are trying to study. Mean plus or minus error bar. bam.

Oh, they are getting warmed up now. A little worrisom on the joker rate but maybe people were just fired up to click the first option? Anyway, CLEARLY, a correlation can exist without ever point being perfectly predictive of the central tendency of the relationship between two variables. And CLEARLY the fact that there is some variability in one measure does not mean it does not tell us, on average, about the other measure if those two things are correlated. And as good scientists who are able to understand the idea of central tendency and error, we do not throw out a correlation if it does not form an invariant line. Or at least most of us do not.

Now, what about the simplest of experimental designs? The two factor, two level quad box that appears in the first chapter of any Experimental Design text book?

Well OF COURSE good scientists understand that there can be not just random variation in a measure, there can be non-random variation. I.e., an influence of another factor! And this may be a constant but is most often a variable influence. Which, gasp, may INTERACT with the first factor in some way….often a variable way. So of course these good scientists, many who deal with this very simple reality of the natural world on a daily basis, report that they are well aware of such things and would never toss a measure just because it was influenced in an identifiable way by more than factor. Geez, don’t insult our intelligence here.

Okay, so we finally got to the water drinking part.

Good god, scientists let all of their training go right straight out of the window when it comes to the business of being a professional scientist.

2 Responses to “You can lead a horse to water but…”

  1. Draino Says:

    …but the horse won’t drink if the water looks politically incorrect. So why did you lead the twitter horses to a non-PC water source, DM? What’s your point?


  2. drugmonkey Says:

    PC is not a thing. try again.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: