Well well. My fallse discovery rate paper has passed 50,000 full text views (and 6.600 downloads). Love #openaccess http://t.co/lFsDIx7TqI
Well well. My fallse discovery rate paper has passed 50,000 full text views (and 6.600 downloads). Love #openaccess http://t.co/lFsDIx7TqI
Well well. My fallse discovery rate paper has passed 50,000 full text views (and 6.600 downloads). Love #openaccess http://t.co/lFsDIx7TqI
Well well. My fallse discovery rate paper has passed 50,000 full text views (and 6.600 downloads). Love #openaccess http://t.co/lFsDIx7TqI
Well well. My fallse discovery rate paper has passed 50,000 full text views (and 6.600 downloads). Love #openaccess http://t.co/lFsDIx7TqI
Well well. My fallse discovery rate paper has passed 50,000 full text views (and 6.600 downloads). Love #openaccess http://t.co/lFsDIx7TqI
Well well. My fallse discovery rate paper has passed 50,000 full text views (and 6.600 downloads). Love #openaccess http://t.co/lFsDIx7TqI
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
@AlxEtz Will you blog about false discovery rates? eg http://t.co/X1NEc1GDCR
Want to make a fool of yourself 30% of the time? Then use p = 0.05 to support your results... http://t.co/oHp9VjhHvb
Want to make a fool of yourself 30% of the time? Then use p = 0.05 to support your results... http://t.co/oHp9VjhHvb
Observation of a p∼0.05 means nothing more than ‘worth another look’ http://t.co/JJJHWKSU4S
Colquhoun: An investigation of the false discovery rate and the misinterpretation of p-values: http://t.co/tHyhjciHcg
"The standard approach in teaching [of p-values] has simply been an abysmal failure" http://t.co/JJJHWKSU4S
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
The function of significance tests is to prevent you from making a fool of yourself. http://t.co/6bnNOy6Jdy by @david_colquhoun
The beauty of p-values: "If, as is often the case, experiments are underpowered, you will be wrong most of the time." http://t.co/JJJHWKSU4S
Using p=0.05 to suggest you made a discovery? You're wrong at least 30% of time. Insightful paper http://t.co/90GjCi1sjI via @TahaYasseri
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
@rafaelprietoc what's the upcoming stats piece about? Wondered about a piece on http://t.co/X1NEc1p2Lj
@maelanfontes @puddleg @PolderPaleo @OscarPicazo very useful http://t.co/222C3Tc1RV by @david_colquhoun.changed my interpretation of results
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. CRAZYY! @TahaYasseri @david_colquhoun http://t.co/gEpm0q6sUC
@MartinNutrition @TStellingwerff not sure why http://t.co/222C3Tc1RV is e.g. of #lchf worsening sprinting? http://t.co/478wgAz0DD
@maelanfontes @puddleg @PolderPaleo @OscarPicazo very useful http://t.co/222C3Tc1RV by @david_colquhoun.changed my interpretation of results
@rafaelprietoc what's the upcoming stats piece about? Wondered about a piece on http://t.co/X1NEc1p2Lj
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Using p=0.05 to suggest you made a discovery? You're wrong at least 30% of time. Insightful paper http://t.co/90GjCi1sjI via @TahaYasseri
Using p=0.05 to suggest you made a discovery? You're wrong at least 30% of time. Insightful paper http://t.co/90GjCi1sjI via @TahaYasseri
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
"An investigation of the false discovery rate and the misinterpretation of p-values" http://t.co/wpUT2wotXT
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Using p=0.05 to suggest you made a discovery? You're wrong at least 30% of time. Insightful paper http://t.co/90GjCi1sjI via @TahaYasseri
Using p=0.05 to suggest you made a discovery? You're wrong at least 30% of time. Insightful paper http://t.co/90GjCi1sjI via @TahaYasseri
"And never use the word ‘significant’" http://t.co/7NezKMVm74 #statistics
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
Used p=0.05 to suggest you made a discovery: you're wrong at least 30% of time. Great paper by @david_colquhoun http://t.co/huuuqcxilA
.@KyleCranmer @altmetric Well the altmetric score of 674 on http://t.co/X1NEc1p2Lj is embarrassing, Shows I spend too much time on twitter
"If you use p=0.05 to suggest that you have made a discovery, you will be wrong at least 30% of the time." http://t.co/Nfnf8I1D9E
An investigation of the false discovery rate and the misinterpretation of p-values http://t.co/YknyARnB6w
.@KyleCranmer @altmetric Well the altmetric score of 674 on http://t.co/X1NEc1p2Lj is embarrassing, Shows I spend too much time on twitter
@melb4886 or most recently, RS Open Science -hasn't even got an IF yet, but what the hell http://t.co/lFsDIx7TqI
@snim2 well my advice is give P value, descrobe sensibly (never say sig or non-sig) http://t.co/Hb7so5LL6d and show effect size as mean + CI
@snim2 yes -full paper http://t.co/X1NEc1p2Lj Simpler video http://t.co/jbEqYe5N1q Still simpler blog http://t.co/oCm0RwfBio
@snim2 yes -full paper http://t.co/X1NEc1p2Lj Simpler video http://t.co/jbEqYe5N1q Still simpler blog http://t.co/oCm0RwfBio
@snim2 yes -full paper http://t.co/X1NEc1p2Lj Simpler video http://t.co/jbEqYe5N1q Still simpler blog http://t.co/oCm0RwfBio
An investigation of the false discovery rate and the misinterpretation of p-values http://t.co/7b4woNarNV
.@DavidPriceUCL in fact I prefer typography in home-made version in arXiv http://t.co/relFUWcK7f to that in RS's pdf http://t.co/X1NEc1p2Lj
use p=0.05 to claim you have discovered an effect? then >30% risk of looking foolish Read this http://t.co/Hmss5SdeCp Bravo @david_colquhoun
@omaclaren of course but I made assumptions perfectly clear in http://t.co/X1NEc1p2Lj They are the standard ones @learnfromerror
why is p=0.05 the default & often only option considered? min false discovery rate using p values ~30%. http://t.co/A7rI1Ntfgv
"If you use p=0.05 to suggest that you have made a discovery, you will be wrong at least 30% of the time" http://t.co/5LhtOCB7yQ #statistics
Residents finishing research projects: beware the "significant" p-value. http://t.co/fJkngBVP0h h/t @DavidJuurlink
Residents finishing research projects: beware the "significant" p-value. http://t.co/fJkngBVP0h h/t @DavidJuurlink
¡Bueno! http://t.co/u7YhwbJ2hO
Enlightening stuff from @david_colquhoun. False discovery rates and misinterpretation of the P value. http://t.co/5RipkU2aAr
An investigation of the false discovery rate and the misinterpretation of p-values by @david_colquhoun http://t.co/ngXWfYfjHj
An investigation of the false discovery rate and the misinterpretation of p-values by @david_colquhoun http://t.co/ngXWfYfjHj
@TimBrock_DtD @HannaIsotalus @wandedob R is fine for stats and not too hard - took 3 days to earn enough for http://t.co/X1NEc1p2Lj
@TimBrock_DtD @HannaIsotalus @wandedob R is fine for stats and not too hard - took 3 days to earn enough for http://t.co/X1NEc1p2Lj
Enlightening stuff from @david_colquhoun. False discovery rates and misinterpretation of the P value. http://t.co/5RipkU2aAr
H/T @LeighBreen: http://t.co/FewDDs2KYg @david_colquhoun #sotrue (More academics and researchers should read this...) http://t.co/8Q6sRovuEq
Enlightening stuff from @david_colquhoun. False discovery rates and misinterpretation of the P value. http://t.co/5RipkU2aAr
Blimey -my not-very-original paper on false discovery rate has passed 6,000 pdf downloads and 45,000 full text views http://t.co/X1NEc1p2Lj
Blimey -my not-very-original paper on false discovery rate has passed 6,000 pdf downloads and 45,000 full text views http://t.co/X1NEc1p2Lj
Enlightening stuff from @david_colquhoun. False discovery rates and misinterpretation of the P value. http://t.co/5RipkU2aAr
Blimey -my not-very-original paper on false discovery rate has passed 6,000 pdf downloads and 45,000 full text views http://t.co/X1NEc1p2Lj
Blimey -my not-very-original paper on false discovery rate has passed 6,000 pdf downloads and 45,000 full text views http://t.co/X1NEc1p2Lj
Blimey -my not-very-original paper on false discovery rate has passed 6,000 pdf downloads and 45,000 full text views http://t.co/X1NEc1p2Lj
Blimey -my not-very-original paper on false discovery rate has passed 6,000 pdf downloads and 45,000 full text views http://t.co/X1NEc1p2Lj
@mahkusjaybee @jameswilsdon I wouldn't disagree with that aspect! http://t.co/X1NEc1p2Lj
@rblotnicky @itsmepanda1 @EDavis32 @sillyunclesteve @YoniFreedhoff And http://t.co/cRpIz3xxWY The gist is, small sample sizes mislead.
@lakens The frequentist p tests whether a finding depends on sampling, it does not test the effect (i.e. m=zero) http://t.co/oOMGh1m1ov
Why ‘statistically significant’ and p ≤ 0·05 doesn't actually mean anything useful: http://t.co/a4Kycqo5WO
Why ‘statistically significant’ and p ≤ 0·05 doesn't actually mean anything useful: http://t.co/a4Kycqo5WO
FDR es parte de la solución http://t.co/teA2rvDrZm
use p=0.05 to claim you have discovered an effect? then >30% risk of looking foolish Read this http://t.co/Hmss5SdeCp Bravo @david_colquhoun
use p=0.05 to claim you have discovered an effect? then >30% risk of looking foolish Read this http://t.co/Hmss5SdeCp Bravo @david_colquhoun
use p=0.05 to claim you have discovered an effect? then >30% risk of looking foolish Read this http://t.co/Hmss5SdeCp Bravo @david_colquhoun
use p=0.05 to claim you have discovered an effect? then >30% risk of looking foolish Read this http://t.co/Hmss5SdeCp Bravo @david_colquhoun
@analysis_factor I think confidence intervals are really informative though. Have you seen this ? http://t.co/RWufvyyL0I p=.001 the new.05
An investigation of false discovery rate and misinterpretation of p-value. http://t.co/GA48yJM2Nu v. @royalsociety @statsmethods @statsepi
An investigation of false discovery rate and misinterpretation of p-value. http://t.co/GA48yJM2Nu v. @royalsociety @statsmethods @statsepi
@stephensenn the fact is that if you use the bog standard assumptions of the t test, P values have a problem http://t.co/X1NEc1p2Lj