Conversation

Notices

1. D00B ⛄ :blobcatsanta: (d00b@seal.cafe)'s status on Friday, 30-Dec-2022 12:11:17 JST D00B ⛄ :blobcatsanta:

   • bot :kiwi_dumbbell:
   • Parker Banks
   • Nathanael Nerode
   @parker @bot @neroden leftyfags LOVE their studies. But they don't know over 90% of all studies can't be proven after print.
   • Parker Banks (parker@pl.psion.co)'s status on Friday, 30-Dec-2022 12:11:15 JST Parker Banks

     in reply to

     • bot :kiwi_dumbbell:
     • Nathanael Nerode
     @D00B @bot @neroden usually for pharmaceuticals they have a drug that sucks, conduct thousands of tests till they find one measurement that "significantly" differs from controls, then publish their results as if the non-significant tests never existed
   • D00B ⛄ :blobcatsanta: (d00b@seal.cafe)'s status on Friday, 30-Dec-2022 12:11:16 JST D00B ⛄ :blobcatsanta:

     in reply to

     • bot :kiwi_dumbbell:
     • Parker Banks
     • Nathanael Nerode
     @parker @bot @neroden yeah isn't it wacky how it goes
     >we have this retarded drug that sucks
     >gib money to retards in labcoats
     >they do study writing the end first
     >they review their own study under a different name
     >publish
     >FDA niggers use it to push new shit drug
     rinse and repeat.
   • Parker Banks (parker@pl.psion.co)'s status on Friday, 30-Dec-2022 12:11:17 JST Parker Banks

     in reply to

     • bot :kiwi_dumbbell:
     • Nathanael Nerode
     @D00B @bot @neroden between no one publishing replication studies because they get no funding, the file drawer problem, and every paper having hundreds of statistical tests performed before arriving at the published results, it's way worse than p<.05. but at least the above "debunked" paper is a meta analysis of 109 different studies.