[alexander_busek]

Instead, in its most fully specified model, it [the study] reports that 99.6839% of black babies born with a black attending physician survived compared with 99.5549% of black babies born with white attending physicians, a difference of 0.129%.

100% - 99.6839% = 0.3161%.

100% - 99.5549% = 0.4451%.

Meaning: When a Black physician is in attendance, only 0.3161% die.

And when a White physician is in attendance, 0.4451% die.

0.4451% divided by 0.3161% = approx. 1.41.

Meaning: The death rate with a White physician is about 41% higher.

HOWEVER, we are talking about a comparison of two tiny slivers of data-sets.

I would hazard the claim that:

a. This is in the range of uncertainty and/or the data-sets are too small to be statistically significant, or

b. A multi-variant analysis taking into consideration ALL factors would cause the difference to vanish.

FURTHER: What, no Asian or Latino attending physicians?! Don't they exist?!

Regards,

[Samurai_Jack]

0.4451% divided by 0.3161% = approx. 1.41.

Meaning: The death rate with a White physician is about 41% higher.

Thanks for this perspective post. So noise from the left would be that kentangie was not lying, she just didnt tell the whole truth. Which is similar to the lie they told about the 2020 election. Justifying it by saying it was ‘rigged’ and not ‘stolen’. But the difference between those two is pedantic at best.

[sonjay]

100% - 99.6839% = 0.3161%.
100% - 99.5549% = 0.4451%.
Meaning: When a Black physician is in attendance, only 0.3161% die.
And when a White physician is in attendance, 0.4451% die.
0.4451% divided by 0.3161% = approx. 1.41.
Meaning: The death rate with a White physician is about 41% higher.

That's not how it works. To compare the death rates, you would take the difference between the two rates and divide that by the rate against which you are comparing the other one.

For example, in simpler terms: a 60% death rate (group A) vs a 40% death rate (group B). The difference is 20. 20 divided by 60 = .08. So the death rate for group A is 33% higher than that for group B.

With these black babies, the difference is .00129409: 99.4451 - 99.3161 = 0.129. Then 0.129 divided by 99.3161 = 0.00129888. In percentage terms, that's 0.1% higher. NOT 1.41% higher.

[Svartalfiar]

Or look at it from the other direction.

99.6839% - 99.5549% = 0.1290%

0.1290 / 99.5549 = 0.0012958

So the survival rate with a black doc is only 1.0013x higher than with a white doc.



I'm gonna go with your points a) and b) likely both being correct. A sample size of only 775 babies will result in one baby being that 0.129% difference.

[fuzzylogic]

I always find that the use of “simple stats” is nothing but a tool to distort truth, things are usually more complex. Same with the “wage gap”.

I could see a study with some numbers like:

Block doctors: 1 in 1000 black babies die
White doctors: 2 in 1000 black babies die

So, “twice as many babies died in the hands of white doctors vs. black doctors”. It would be an absurd basis for anything though, as this is a single study with a limited sample size and would fall into the margin of error & is statistical noise.

But that is what the left always does with “simple stats”. Anytime you hear them, just scratch the surface, just a little, and it falls apart. Remember the Roe vs Wade argument? That “back alley abortions kill xxxx women per year!”...yeah, they admit they just pulled the number out of their a**.

It’s disturbing when a SC justice does it though, it’ll be repeated ad-nauseam.