From Steve Kirsch:
Summary of the approach used by Professor Switkay
In my initial analysis of the survey data, I looked at the odds at the extreme points (no vax vs. fully vaxxed) to compute an odds ratio.
What Professor Switkay did was to look at all the intermediate data points in the survey and check for a dose-response relationship that would be consistent with a causality hypothesis.
So instead of looking at two data points per condition (the odds for unvaxxed and vaxxed), he looked at all 5 data points (since there were 5 different vax levels in the survey: no, low, medium, high, very high) and then fit a line through them.
For each condition, he did a regression analysis on the log of the odds and computed the value for the Pearson correlation coefficient (aka “r”), t statistic, slope of the line through the points (an indicator of the effect size), and more.
If vaccines are causing a condition in a linear fashion, plotting the log odds should be a straight line. In short, if the log of the odds is a straight line, it means that if you double the dose, you double the response.
He found that indeed, the log odds lined up in a straight line and the “fit” of the line to the points was amazing for many of the conditions. An r value of .97 for example is something you rarely see in real world data. It’s basically “nearly perfectly correlated.”
Depression had r=.99 and t statistic=12.4.
Sexual orientation issues had r=.97 and t statistic=7.4.
These are simply stunning effects that didn’t happen by chance; they happen because vaccines are causing them.