Today's paper that I'm checking is not ivermectin, another treatment for COVID-19, and thus far every single percentage I've looked at in every table is impossible Cited 18 times so far
-
-
The whole thing is just...very odd. The p-values are also bizarre - I get a chi-squared p-value of 0.0005, not 0.036, for the last line
Show this thread -
For those wondering what I mean when I say "impossible"https://twitter.com/GidMK/status/1428174111834923018?s=20 …
Show this thread -
Worth noting that lots of papers have one or two impossible values - often there is a person missing from one line in a table, that throws off the percentages. It's not ideal, but it's a very simple mistake It becomes worrying when this mistake is repeated a lot
Show this thread
End of conversation
New conversation -
-
-
Table 1 is definitely swapped, at least. If we recalculate the % considering 44 to the treated arm and 43 people in the placebo data seems to match the %. This is a veeeeeeeery criticized paper here in Brazil and is also considered the new silver bullet for the gov. supporters
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Yeah. Swapping the N=43 vs 44 would fix some of it. I notice that 26% more had recovered by day 1, and the difference in recovery % is lower every day after that. Seems a bit odd to me.
-
Yep, they've crossed over the n=43 and 44 numerous times in the paper quite clearly, but there are plenty of percentages that are impossible even with 43 or 44 so it's all very confusing
- Show replies
New conversation -
-
-
They say in the text that they only tested 2/3 of the participants of the study, because of a low budget. Maybe that’s why the percentages are off.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.