Carl Fried

Statistician to the stars

[Frank Harrell @f2harrell]

"Real world data" is an almost meaningless phrase. Our new @bmj_latest paper pushes the more helpful phrase "routinely collected data" and explains the many limitations of such for research purposes. https://t.co/VvsxLQvhfk #Statistics

[Carl Fried @rnorm01]

@Cacciamani_MD @EvidenceOpen ?

[Carl Fried @rnorm01]

@whitfieldlewis6 @ZKForTre Start here. Thanks https://t.co/TaOhXGY5As

[Carl Fried @rnorm01]

@whitfieldlewis6 I recommend you, and others, follow people who know what they are talking about https://t.co/bSqdJxwzcU

[Carl Fried @rnorm01]

@whitfieldlewis6 I recommend you, and others, follow people who know what they are talking about https://t.co/bSqdJxwzcU

[Richard Riley (R²) @Richard_D_Riley]

Periodic reminder for meta-analysis folk:

I-squared is as an estimate of between-study heterogeneity ❌

I-squared measures the proportion of total variability due to between-study heterogeneity ✅

https://t.co/Uha3OFrKeQ

[Carl Fried @rnorm01]

@whitfieldlewis6 @ZKForTre "It’s a measurement of heterogeneity across studies in a meta analysis". Why don't *you* Google it Dr. Dunning-Kruger?

[Carl Fried @rnorm01]

@whitfieldlewis6 @ZKForTre for a start, why don't you explain what I^2 has to do with heterogeneity?

[Carl Fried @rnorm01]

@ArnavAgarwalMD @JClinEpi @UofAResearch @UAlberta_DoM @UAlberta_FoMD @CVC_UAlberta @GuyattGH With you on CI crossing the null not being evidence for no effect. But since when has a point estimate being > or < 0 been evidence of benefit or harm?

[NonsparseOncologist @5_utr]

🧵 Pandya et al. just landed in Cell with a “14-protein signature that predicts lung cancer 5 years out.” Nature did a news piece. The press is losing its mind. I read the methods. Someone has to. 1/10

[NonsparseOncologist @5_utr]

If you think putting 107 biased observational studies in a blender yields the truth, @bmj_latest has a new paper for you. Wang et al. try to validate Target Trial Emulations against RCTs, but commit basic biostatistical sins that completely mask individual model failures. 🧵

[Rebecca Shatsky, MD @Dr_RShatsky]

I am not loving this trend of AI generated infographic explanation of pivotal clinical trial data. Many of them are straight up wrong as they aren’t peer reviewed or even created by an expert in that cancer type. Be CAREFUL when using these to learn about data and look at the source. #bcsm #breastcancer

[Carl Fried @rnorm01]

@SyedAAhmad5 Read the figure legend then delete the tweet

[Carl Fried @rnorm01]

@kevinnbass Perhaps read the figure legend. This is like a graph of probability of failing to complete med school but the sample is Kevin Bass

[Carl Fried @rnorm01]

@DrDiGiorgio "cummulative proportion retained". Probably wouldn't take this too seriously

[Carl Fried @rnorm01]

@RicBertolo @jryckman3 If only there were a strong relationship between publishing in prestigious journals and expertise

[Carl Fried @rnorm01]

@5_utr Yeah but the impact factor is 43, so there!

[Carl Fried @rnorm01]

@GavittWoodard @GraceHaMD @nchudgar @BrendonStilesMD @MontefioreNYC @MarcV_MD @TamarNobel @huroqu90 @AATSHQ @ReneHPetersen @olgaostrovetsky what's the E value?

[Carl Fried @rnorm01]

@yudapearl examples? Presumably easy if it occurs so often

[Carl Fried @rnorm01]

@WhiteHouse now show the median

[Carl Fried @rnorm01]

@CalvinKLai reminds me of