Hammad Sheikh, PhD 🔍

All bad takes are mine alone; but the good ones are also from @UniofOxford @KingsCollegeLondon

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

Very cool idea for quant folks: When you plan a study, do a power analysis for an equivalence test. Then if the effect you are looking for is not significant, you will be able to “establish” a null effect. Either way you win!

[Daniël Lakens @lakens]

Reminder that you can not conclude the absence of an effect based on p > 0.05. You need equivalence tests instead. When planning your study, include a power analysis for an equivalence test to make sure your study is informative - also if you find a null result.

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

@Lobber0 @keysmashbandit They don’t.

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

Haven’t posted here for a while. Does anyone know if sharing that I am also on blusky at https://t.co/baHTaQrhBi will get you banned? Asking for a very close friend.

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

@kareem_carr Yes, this is not how the force works.

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

@PaulESpector @lakens Hmm. Are the assumptions and stats more complex in Bayesian than in frequentist approaches? Could you elaborate?

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

@isaac_flath @guitchounts @GaryMarcus It’s almost like circular reasoning: Hard unsolved problems are hard to solve with current approaches…

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

@social_brains The casual anti-semitism one encounters here is mind blowing. In many places, it is part of the vernacular. And yet, many Jews and Israelis decided to settle here and seem happy. You may be right, this may be the “top of the mountain.” :-/

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

@C_Kavanagh You are right. The comparison is exceedingly silly. Looks like a talking point to me, not a serious analysis.

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

@SpencrGreenberg I added also: "Speak like Yoda"

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

@ajordannafa It’s a big step. And it will feel amazing to have this behind you.

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

@kareem_carr Blerd! ;-)

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

@Mark_Zheng95 @Oliver_S_Curry How strong is the evidence for such a strong claim?

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

@MikeyPasek Yep. Had to take the TOEFL, too. Got a perfect score, having lived and studied in Ireland, published papers in English, and in general grown up with English TV... It’s ridiculous.

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

@ajordannafa So professorial!

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

@rmkubinec Is the statement supposed to imply that the statement cannot be trusted because it’s author has financial incentives to discourage an academic career? Or is it more about that people with vested interest cannot be trusted with their statements?

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

@jayvanbavel So, instead of charging for posting, maybe, social media platforms should charge for posting misinformation?

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

@jayvanbavel @AlexandraWorm @AdamBCohen1 You are very precise in your summary. Just want to point out the risk of the ecological fallacy to the readers. This is state-level data, on the individual level, the relationship can be totally different, when reversed.

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

@ajordannafa Thank god for the “unsend” button in gmail. Has saved me much embarrassment.

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

@jonaslindeloev @franbetteo Any idea of the data generating process. Some domain knowledge can do wonders for modeling.

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

@HilaryMatfess Get a birth doula if you can afford it. Doctors/nurses will be pushy and may pressure you into bad decisions. Happened to us and we sent days in the NICU because of an infection as a result of unnecessary intervention.

[Hammad Sheikh, PhD 🔍 @HammadSheikh]

@MohammadAtari90 @UMassAmherst @PBSUmass One day, you will branch out to research on it with gaming, and then the time will be ripe.