-
-
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
But also an error in coding the missing data too. Why should we expect physicians to naturally be able to do statistics? Would we expect a statistician to take out a gall bladder?
-
I’m not sure the error was by physicians. But patients would expect anyone doing data analysis that influences treatment to take the same care expected during a gallbladder operation.
Kraj razgovora
Novi razgovor -
-
-
I suspect it is due to factor to numeric conversion or due to undefined factor level
I used to make crosstables whenever I recode a variable. Do you suggest any double check step? -
Who knows where it went wrong here. Plotting raw data helps. Publishing code without data doesn’t help. All recoding should be checked in simple table. In R this is just `table(old_var, new_var)`. Missing data imputing error just seems a mistake.
@harryclax@dnepo@themalariaarea - Još 2 druga odgovora
Novi razgovor -
-
-
How embarrassing for them. But did they also have to revise their theories and assumptions? Data can be wrong but confirmatory bias is powerful.
-
The paper was republished switching the interpretationhttps://jamanetwork.com/journals/jama/fullarticle/2752467 …
- Još 1 odgovor
Novi razgovor -
-
-
Code & data should be part of peer-review. There should be at least one specialist reviewer with data & stats skills.
-
I agree with the sentiment, but it is often difficult to share patient-level data in this way. The analysis was in Stata, there may not be any ‘code’? We share analyses using shinyfit, but this wouldn’t help catch coding errors.https://github.com/ewenharrison/shinyfit …
- Još 3 druga odgovora
Novi razgovor -
-
-
@JAMAOnc@JAMA_current I appreciate the honesty shown by the authors in today's time when people do anything to get a paper published in a good journal.. -
Absolutely ! Mistakes can happen. Identifying and correcting them is not smth to be ashamed of. On the contrary.
- Još 2 druga odgovora
Novi razgovor -
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.