The article doesn't mention "fudged" data
What data was "fudged" ?
👀
In data analytics and statistics, "fudging data" refers to intentionally altering, inventing, or selectively omitting data points to achieve a desired result. While often done to make reports look better or to fit a specific hypothesis, this practice is scientifically and ethically unacceptable.
Common Types of Data Fudging
"Fabrication: Creating entirely fake data out of thin air to fill a gap or fabricate results.
*Falsification (Cherry-Picking): Only using data that supports your claim while ignoring or hiding valid data that contradicts it.
*P-hacking (Data Dredging): Rerunning tests or tweaking variables until a dataset finally yields a statistically significant \(p\)-value.
*Overfitting: Excessively tuning a machine learning model to historical data so it fails entirely on new, real-world data.
...
Alito relied on a flawed / fudged study from the Trump DoJ to assert that voter turnout among blacks now equals or exceeds that of whites. This supports his view that affirmative action programs are no longer needed.
The data is fudged or flawed because the model is flawed. The study used voting age population to determine turnout. This includes those ineligible to vote. The preferred models use citizen voting age,, or better yet, registered voters..
It turns out, for whatever reason, using voting age population is not only dubious, it also deflates white turnout quite a bit more than black turnout..
The more accepted methods, using citizen voting age or registered voters, both show black turnout declining in relation to white turnout.
Omitting the studies using the more accepted models looks like intentional cherry picking or fudging rather than an unintended flaw..
Recentwtf is wrong with the rig... by illiniray
No New Wars! Well, maybe ... by ThePAMan
In case you haven't figur... by Reacher
Total Members: 142