| Ok, I should have read further and posted a full reply. I had seen his data before, and it looked a bit scary, so I am interested, but I also quickly noticed that it lacked statistical error bars and was excluding any data that contradicted his hypothesis. Would it change your perspective if you learned that the still birth rate in 2022 was still lower than every previous year, except for the years 2019 and 2020 that he uses as comparison? https://www.ons.gov.uk/peoplepopulationandcommunity/birthsde... Does it change your perspective if you learned that the birth rate has been dropping about 2-3% per year, except for the year in comparison (2021) when it was 2% higher than the previous year? https://www.ons.gov.uk/peoplepopulationandcommunity/birthsde... Does it change your perspective if you learned that vaccinated mothers appear to have been slightly more likely to have a healthy, normal birth over this same time period (fig 3-9), though the difference was not statistically large? https://assets.publishing.service.gov.uk/government/uploads/... All of these links are directly from his blog post, so I know he also had access to this data, which shows it was safe and even beneficial to you to get vaccinated, and he apparently chose to ignore it. > even if it is temporary after one shot If it doesn't reoccur though, it is temporary, regardless of what else happens. But the data would already have included the effect of boosters by now. In some ways, the blog post also suffers from a lesser known problem of being too-big of an effect. It was >8% decrease when <20% of the population had a vaccine, so now that >60% has had a vaccine, we should see a >25% drop in births. That is too big an estimated effect he's observing in one specific area for it to be invisible to other statistical measures and countries. So that hints something else is wrong with his analysis, failing to account for something, or taking small numbers as divisors to make them look bigger. I had heard a fun podcast on excess effect size being a possible yellow-flag for statistics, but cannot locate the exact one for you. If I recall right though, it was related to Andreas Glöckner's discussion of the size of the mental depletion effect linked from
https://freakonomics.com/podcast/in-a-job-interview-how-much.... In that article he shows how the correlation was too strong between "mental depletion" and outcome, that it likely points to there not being a causal-link. |
Thus, to come back to belltaco's point, his post shows there is far more evidence for invisible space aliens (75% unexplained correlation) who only target Hungarians than for vaccines (25% possible correlation).