«

»

Hand-Eye Theater

CAUTION: THIS POST CONTAINS MATH

AUG 23, 2023 — It’s becoming more and more apparent that the NCSBE has designed a hand-eye process to fool the public into believing they have the ability to detect any real or imagined machine manipulation.

One clear indication of said manipulation would involve making sure the marked ballots match up with the posted results. Thankfully, NC law requires them to do that . . . kinda.

Statistical Significance

I warned you up front, but let’s start with a layman’s definition of the term from the Harvard Business Review:

“Statistical significance helps quantify whether a result is likely due to chance or to some factor of interest . . . . When a finding is significant, it simply means you can feel confident that’s it real, not that you just got lucky (or unlucky) in choosing the sample.”  — Amy Gallo, Harvard Business Review

In the election business, “error” means the ballot does not tabulate the way a voter intended.

North Carolina law is usually written and vetted by attorneys who never took a graduate-level research methods course.

Despite that limitation they did include this very important gem in NCGS 163-182.1(b)(1): The size of the sample of each category shall be chosen to produce a statistically significant result and shall be chosen after consultation with a statistician.”

We’d love to interview their statistician, but we’re not holding our breaths on that one.

Thanks to the diligent work of NC Audit Force, we’ve learned that in 2022, the percentage of ballots Wake County sampled was a paltry 0.39% and Guilford counted 0.62%. Meanwhile, Gates and Washington counties hand counted over 54% of their ballots sampled. (You may download their 2022 numbers here.)

The 2020 numbers were equally horrifying and I have more details. After their 2020 “audit,” the SBE reported a hand-eye recount of just 387 Mecklenburg County ballots, out of 569,499 cast and Wake counted 1,514 out of 634,423 cast.

We’re not sure what grad school would give a passing grade to a student claiming “statistical significance” with sample sizes that small, but we’d love to see their math.

Here’s ours: Using a sample-size calculator from Calculator.net, Instead of 387 ballots, Meck would have needed to sample 9,442 ballots in order to achieve a 95% confidence interval at a margin of error of 1.

The following chart, from NC Audit Force, shows how the state’s largest “blue” or Democrat-dominant counties grossly fell below any sort of significant hand-eye recount numbers and Republican or “red” counties sampled a far greater percentage of their ballots.

The thin line going through the graph reflects the percentages shown down the right side of the table. In other words, Meck sampled less than 0.20% of their ballots, while Rowan sampled nearly 1.0% of theirs.

Thankfully, sub-paragraph (b) in NCGS 163-182.1 says, “the State Board shall publish in the North Carolina Register the procedures and standards and any changes to them after adoption, with that publication noted as information helpful to the public . . . . Copies of those procedures and standards shall be made available to the public upon request or otherwise by the State Board.”

To that end, we’ve asked the SBE to show us the guidance they sent out to the County Boards of Elections on this matter.

We hope to update this story or even post a new one once we see how CBEs are supposed to do that statutory process.

Either way, lawmakers need to ask some hard questions.