Conversation
Notices
-
Embed this notice
jews rape kids
-
Embed this notice
BTW @jeffcliff this is also a good example of the danger in trusting p-values. You can't get better than p < 8e-17, but the p-value is meaningless because the null hypothesis used in the 'study' fails to account for obvious data artifacts. One wonders how many COVID policy studies rely on similarly flawed analyses. In my opinion, you cannot trust a p-value unless you thoroughly understand the data-collection procedure
-
Embed this notice
@SuperSnekFriend @elftummy This anon is retarded. I took his filtered data set 'missing_kids_data_2000_2024.csv' and plotted the number of cases per month (pic 1). The number of cases explodes in recent years likely due to improved data gathering. The data for 2024 only goes up to May 2024 because that is present-day. This arbitrary cutoff implies that the months leading up to May will have the largest number of cases (pic 2), because only those months see the 2024 data.
Passover always falls in March or April. Therefore, you would expect - just from the data cutoff artifact (see pic 2) - that there are roughly twice as many cases in a Passover window compared to a window of the same size in a random month (possibly June or later). This fully explains what is shown in the OP's plot.
cc @ceo_of_monoeye_dating @jeffcliff
-
Embed this notice
@elftummy Link to repo and higher res of the graphs:
anonymous.4open.science/r/interesting_data-06F9/README.md