RT @Winthernet: New study shows "most common software packages for fMRI analysis can result in false-positive rates of up to 70%" - https:/…
RT @sjcockell: 2/ is anyone really surprised by the latest fMRI snafu? Fast and loose stats used to sell 'story' https://t.co/5bdlFVZzyS
And that is why one should keep data stored for more than 10yrs. https://t.co/ZDjCcG0GCU #falsepositive
https://t.co/5TXKLFdFP8 "... as 3dClustSim is one of the most popular choices for multiple-comparisons correction." I wonder why...
this seems like it would be a problem if any cog. neuroscientists actually cared about stats https://t.co/Oe6d9Z2HwK https://t.co/7o02Nrfra9
RT @russpoldrack: Cluster failure: Why fMRI inferences for spatial extent have inflated false-positive rates https://t.co/69oIl7wFHN by @te…
New study shows "most common software packages for fMRI analysis can result in false-positive rates of up to 70%" - https://t.co/C6wbN5TJ8N
"we found that the most common software packages for fMRI analysis can result in false-positive rates of up to 70%": https://t.co/c1ICT9Ty3g
“Cluster failure: Why fMRI inferences for spatial extent have inflated false-positive rates” https://t.co/Dp214i129l https://t.co/DGhjijylp4
Stunning study: "most common software packages for fMRI analysis can result in false-positive rates of up to 70%" https://t.co/ZEFGaYB883
Cluster failure: Why fMRI inferences for spatial extent have inflated false-positive rates - https://t.co/j3q8L5z3xY
Cluster failure: Why fMRI inferences for spatial extent have inflated false-positive rates https://t.co/as4Wm5JMuu
Interesting to see how this plays out: 40,000 (or perhaps only 16,500) fMRI studies questioned https://t.co/Lz2zy1DBTW #fMRI #statistics
RT @katecrawford: Here's a link to the original journal article: "These results question the validity of some 40,000 fMRI studies." https:/…
RT @katecrawford: Here's a link to the original journal article: "These results question the validity of some 40,000 fMRI studies." https:/…
FMRI disaster.40,000+ fMRI papers published in last 15 years under scanner https://t.co/aB5bRtBYXD @Neuro_Skeptic #neuroscience @PLOSNeuro
Allegedly, false positives up to 70% due to fMRI software programs, thousands of studies possibly in question. https://t.co/FcIhfeVY2z
RT @russpoldrack: Cluster failure: Why fMRI inferences for spatial extent have inflated false-positive rates https://t.co/69oIl7wFHN by @te…
"We found that the most common software packages for fMRI analysis can result in false-positive rates of up to 70%." https://t.co/G82s0j5f6h
RT @katecrawford: Here's a link to the original journal article: "These results question the validity of some 40,000 fMRI studies." https:/…
RT @katecrawford: Here's a link to the original journal article: "These results question the validity of some 40,000 fMRI studies." https:/…
RT @katecrawford: Here's a link to the original journal article: "These results question the validity of some 40,000 fMRI studies." https:/…
40k studies may be affected: #FunctionalMRI analysis software inflates #brainstudies' results' statist. significance https://t.co/RLd3ngN5sM
ooooo fMRI results. we all sorta know, but brush it under the really big bump under the rug. #neuroscience #fMRI https://t.co/wNNDrA8QzB
Cluster failure: Why fMRI inferences for spatial extent have inflated false-positive rates https://t.co/sW82WL6keb
RT @fMRI_today: Regarding https://t.co/tvKP8khO2W, Tom Nichol’s clarification on how many studies may be actually be influenced. https://t.…
RT @fMRI_today: Regarding https://t.co/tvKP8khO2W, Tom Nichol’s clarification on how many studies may be actually be influenced. https://t.…
A bug in fMRI software could invalidate 15 years of brain research https://t.co/6Q2uYZDqym
RT @quantrad: Disaster and embarrassment from not understanding the algorithms you use. Watch out! https://t.co/ltRwDt2t6I
RT @wandedob: My cluster failure paper with @ten_photos is now finally out in PNAS https://t.co/5DSl5U8Kjv #fMRI #OHBM2016 #openaccess
Paper seems to confirm what many of my peers have been advocating for and noting for years… https://t.co/T3XFApUrVG
RT @fMRI_today: Regarding https://t.co/tvKP8khO2W, Tom Nichol’s clarification on how many studies may be actually be influenced. https://t.…
Remember the type 1 error? Old fMRI software may cause too many false positives. https://t.co/zrlMOIwYbx
RT @fMRI_today: Regarding https://t.co/tvKP8khO2W, Tom Nichol’s clarification on how many studies may be actually be influenced. https://t.…
RT @EdwardTufte: #statistics #datascience #EMR PNAS: fMRI software packages = false-positive rates 70%.Validity 40,000 fMRI studies?https:…
RT @fMRI_today: Regarding https://t.co/tvKP8khO2W, Tom Nichol’s clarification on how many studies may be actually be influenced. https://t.…
RT @fMRI_today: Regarding https://t.co/tvKP8khO2W, Tom Nichol’s clarification on how many studies may be actually be influenced. https://t.…
RT @fMRI_today: Regarding https://t.co/tvKP8khO2W, Tom Nichol’s clarification on how many studies may be actually be influenced. https://t.…
RT @fMRI_today: Regarding https://t.co/tvKP8khO2W, Tom Nichol’s clarification on how many studies may be actually be influenced. https://t.…
RT @republicofmath: fMRI inferences for spatial extent have inflated false-positive rates https://t.co/R6rlshz4Bp https://t.co/oViVxo1PZW
RT @fMRI_today: Regarding https://t.co/tvKP8khO2W, Tom Nichol’s clarification on how many studies may be actually be influenced. https://t.…
RT @fMRI_today: Regarding https://t.co/tvKP8khO2W, Tom Nichol’s clarification on how many studies may be actually be influenced. https://t.…
RT @katecrawford: Here's a link to the original journal article: "These results question the validity of some 40,000 fMRI studies." https:/…
RT @BitesizedMRI: "These results question the validity of some 40,000 fMRI studies" Blimey! https://t.co/3QGtSzCpJo
If you thought your fMRI dissertation idea had already been done, there's good news. https://t.co/wRPwUlRPzl
The most common software pkgs for functional MRI analysis can result in false-pos rates of up to 70%. #radiology https://t.co/huJigjvsis
Wrong software calcs: huge false positive rates in clusterwise fMRI inferences challenges validity of 40,000 studies https://t.co/IbNmXs0kPB
RT @republicofmath: fMRI inferences for spatial extent have inflated false-positive rates https://t.co/R6rlshz4Bp https://t.co/oViVxo1PZW
@Alma_Mahleur @stbmed @Entre2chaises @RadioactiveJib et pourquoi tous le monde en veux à nos machines en ce moment https://t.co/Bu1vwwb8R5 ?
This is huge. Statistical analyses used for thousands of fMRI studies have up to a 70% false positive rate. Yikes… https://t.co/2rBqBCTEX2
"These results question the validity of some 40,000 fMRI studies" Blimey! https://t.co/3QGtSzCpJo
RT @fMRI_today: Regarding https://t.co/tvKP8khO2W, Tom Nichol’s clarification on how many studies may be actually be influenced. https://t.…
RT @fMRI_today: Regarding https://t.co/tvKP8khO2W, Tom Nichol’s clarification on how many studies may be actually be influenced. https://t.…
RT @davidgruber: @katecrawford the associated journal article attached here for added value: https://t.co/fyhwtKw9en
MRI software produces up to 70% false positives. https://t.co/xFFmNKHTIl
RT @fMRI_today: Regarding https://t.co/tvKP8khO2W, Tom Nichol’s clarification on how many studies may be actually be influenced. https://t.…
RT @fMRI_today: Regarding https://t.co/tvKP8khO2W, Tom Nichol’s clarification on how many studies may be actually be influenced. https://t.…
RT @freakonometrics: "These results question the validity of some 40,000 fMRI studies and may have a large impact on the interpretation” ht…
RT @katecrawford: Here's a link to the original journal article: "These results question the validity of some 40,000 fMRI studies." https:/…
@laurieskelly @physicsjackson Check out original: https://t.co/chMQEZ8kYA Big problem seems to be that few neuro papers include raw data
RT @fMRI_today: Regarding https://t.co/tvKP8khO2W, Tom Nichol’s clarification on how many studies may be actually be influenced. https://t.…
Regarding https://t.co/tvKP8khO2W, Tom Nichol’s clarification on how many studies may be actually be influenced. https://t.co/vgzjE1su2K
https://t.co/1eToZHEHRe In theory, we should find 5% false p… https://t.co/G9mjJgkxcG
Interested read, no need to panic but it does raises questions on appropriate controls! https://t.co/nzr3rk88kW
Los pequeños detalles cuando se diseñan algunos modelos en fMRI: resultados reproducibles. https://t.co/RyM9VKnFJm
RT @zooko: Goddammit! Does this mean an entire generation of brain science wasted!? https://t.co/jv0hmOhC0Y https://t.co/Dr4HeRsxAN
RT @emorationality: Popular software for functional #MRI analyses were found to have false positive rates up to 70% https://t.co/5474K14Uz3…
Caution warranted when interpreting fMRI study results due to high false positives, says new article in PNAS: https://t.co/PSfWU0U5zY
RT @emorationality: Popular software for functional #MRI analyses were found to have false positive rates up to 70% https://t.co/5474K14Uz3…
Popular software for functional #MRI analyses were found to have false positive rates up to 70% https://t.co/5474K14Uz3 #ouch #neuroscience
Researchers Say FMRI Software Has Produced "False Positives" for 15 Years via WorkCompAcademy - ... https://t.co/pgvr4yB6gF
RT @katecrawford: Here's a link to the original journal article: "These results question the validity of some 40,000 fMRI studies." https:/…
You wanna know which of those 40000 papers are wrong? Tough luck. That's why we need #opendata and #openscience https://t.co/hWr4vHa9bs
Oha, muss die Geschichte der "wenn Du dies denkst, leuchtet Dein Hirn hier auf"-Versuche neu geschrieben werden? https://t.co/kI3a27PzKL
RT @NetNezva: 40000 papers a lo largo de 20 años basados en resultados erróneos, error de software mediante: https://t.co/yovS8xPGfc
@russpoldrack Your opinion on this issue https://t.co/J3g367hccm
@russpoldrack Your opinion on this issue https://t.co/J3g367hccm
Cluster failure: Why fMRI inferences for spatial extent have inflated false-positive rates from 5 to 70%.... https://t.co/eAfAEPk7ea
RT @sjcockell: 2/ is anyone really surprised by the latest fMRI snafu? Fast and loose stats used to sell 'story' https://t.co/5bdlFVZzyS
2/ is anyone really surprised by the latest fMRI snafu? Fast and loose stats used to sell 'story' https://t.co/5bdlFVZzyS
RT @wandedob: My cluster failure paper with @ten_photos is now finally out in PNAS https://t.co/5DSl5U8Kjv #fMRI #OHBM2016 #openaccess
PNAS - it was always the case that imaging shows statistical tests not images of activity, this highlights the issue https://t.co/S2mKDIwuIQ
Comments/clarifications/corrections from fMRI scientists welcome. 6/n. https://t.co/iQTpQ1fZ35
Implication: problem can be fixed, but will be labor intensive. No need for fMRI to delete its account. 5/n https://t.co/iQTpQ1fZ35
Paper abstract clarifies, problem goes away with non-parametric methods, & doesn’t occur for voxelwise analysis. 4/n https://t.co/iQTpQ1fZ35
Paper abstract clarifies, problem occurs for clusterwise analysis and parametric statistical methods. 3/n https://t.co/iQTpQ1fZ35
Paper significance statement says false positive rate of 70%, not 5%, found using common fMRI software. 2/n https://t.co/iQTpQ1fZ35
Cluster failure: Why fMRI inferences for spatial extent have inflated false-positive rates https://t.co/G4giYwmcbL
Attempting brief corrections to press coverage of PNAS fMRI study. Caveat: have never used fMRI myself. 1/n https://t.co/iQTpQ1fZ35
RT @NetNezva: 40000 papers a lo largo de 20 años basados en resultados erróneos, error de software mediante: https://t.co/yovS8xPGfc
Los análisis más frecuentemente usados en fMRI studies dan un montón de falsos positivos. https://t.co/5P7N4Z9H3f via @MaribelRH
Significant mismatches between press coverage of fMRI bug problem and abstract of study: https://t.co/iQTpQ1fZ35
This is what happens when you only fund science studying “novel” things rather than replicative studies https://t.co/48pzJsl8Ad
RT @NetNezva: 40000 papers a lo largo de 20 años basados en resultados erróneos, error de software mediante: https://t.co/yovS8xPGfc
cluster failure: why fMRI inferences for spatial extent have inflated false-positive rates https://t.co/3nTh68iNow
Wonder what @NeuroscienceDC thinks of this paper: https://t.co/I9J5fNQ2KR Seems to illuminate issues with false positives/hyperbole in fMRI
Según Eklund et al., hasta el 70% de los análisis en fMRI podrían ser falsos positivos. Tsss. https://t.co/oAMLcqJDa9