Lasst uns noch ein bisschen darüber reden, warum an vielen Stellen Studien publiziert werden, die sich am Ende doch nicht replizieren lassen. Denn eigentlich ist Wissenschaft ja daran interessiert, tatsächliche Erkenntnisse über die Welt zu generieren.
RT @realsci_DE: Heißt das, dass man wissenschaftlichen Studien nicht glauben sollte? Nein, das heißt es nicht. Wenn Studien entsprechend…
RT @realsci_DE: Heißt das, dass man wissenschaftlichen Studien nicht glauben sollte? Nein, das heißt es nicht. Wenn Studien entsprechend…
Heißt das, dass man wissenschaftlichen Studien nicht glauben sollte? Nein, das heißt es nicht. Wenn Studien entsprechend der Standards guter wissenschaftlicher Praxis gemacht sind, kann man den Ergebnissen vertrauen.
Überraschend viele Ergebnisse lassen sich nicht replizieren (https://t.co/5qTsHQTBdi, https://t.co/QtMe9kITUY, https://t.co/3wUtyeSGLi). In manchen Forschungsfeldern ist die Realisierbarkeit zwar höher als in anderen, aber insgesamt schaut es trotzdem nich
RT @jayjoseph22: “Selective reporting, selective analysis, and insufficient specification of the conditions necessary or sufficient to obta…
“Selective reporting, selective analysis, and insufficient specification of the conditions necessary or sufficient to obtain the results.” Applies to some of the most famous and frequently cited genetic studies in psychology and psychiatry. https://t.co/g
Japp, reproducerbarhet är så att säga kriteriet på vetenskap. https://t.co/9Q71f13qOg
@raumwurschtler @nicodile Движда се в научни среди и съм се занимавал с научна журналистика и смятам, че съм добре информиран. Но не ми вярвайте. Хвърлете поглед на линка. Или на иследването на Science ( https://t.co/ujMtld4XRO). Шокиран съм, че не сте зап
This began with a discussion of Estimating the Reproducibility of Psychological Science https://t.co/f6pkHVlKYU and this yr, I added The Natural Selection of Bad Science to discuss how the incentive structure in science may select for poor methods https:/
https://t.co/lDhxFk9rfz 100個の心理学実験の再現実験をした話 元論文では97%の実験が有意性のある結果がでたけど、再現実験では36%しか有意性が確認できなかった。でも元論文が間違っていたって言い切れないし、元論文が正しいとも言い切れない
@CompareTheNames Psychology is one of the few disciplines where more than half of their premises don’t replicate in research. https://t.co/VOvgdJy0B8
@dan_s_becker Cute, simple theories published in high impact journals get more scrutiny. Lack of reproducibility and replicability is fairly common in psychology https://t.co/5u9mR2osTj. Fraud is only part of it - bad practices, lack of scrutiny, and bad
@morungos @chemistscorner @stephaniemlee The replication rate for psychology experiments is below 36% https://t.co/uV0OvhESWl
@einstein_ai @bslyBqgExyF897q @sesamethehuman @toxic_cuctus Очевидно, что я взял это отсюда https://t.co/g0ifjv7jZW
Estimating the Reproducibility of Psychological Science (2015) : Open Science Collaboration and others DOI: https://t.co/r97IGSEuKv #decline_effect #meta_analysis #psychology #replication #reproduction_crisis #my_bibtex
Estimating the reproducibility of psychological science | Science https://t.co/oa0XSLEB5D
Hi, good day, how y'all doin? How are my westernised/modernised, cosmetic "psychologists" doin? How does it feel having your memes used against you? :) DOI: https://t.co/kHTGRhZcbu PMID: 26940865 PMID: 31346273.
@nntaleb And it is known they do so "Ninety-seven percent of original studies had statistically significant results. Thirty-six percent of replications had statistically significant results" https://t.co/ELxY8vs9nj
Just about one-third to one-half of the original findings of studies involving experiments that seek to demonstrate a cause/effect relation were observed in a replication study. https://t.co/TsQxiBc3HT
Great opportunity if you want to decide who gets a job, a grant, or just a big pay raise. The only great Science paper for psychology was the one that showed all the other ones are crap. https://t.co/XzYo8nmQJs
Estimating the reproducibility of psychological science https://t.co/MRYZDtk5zv
@BenSlaterNeuro ... and here are two classics from psychology (but indeed the reproducibility crisis is everywhere, not just in psychology): https://t.co/hJoubhu5Bl https://t.co/5eHP41RgPF Well, what is a crisis? Never waste a good crisis. https://t.co/mNi
@MichaelShah @ShriramKMurthi I'd go much further than this essay, but it does say something anyone reading such things should keep strongly in mind: https://t.co/OzkVvpQjjR Perhaps also: https://t.co/bOIJhoBoBm
Tartışma psikoloji alanında yapılan çalışmaların büyük ölçüde tekrarlanamaz olmasının ciddi bir problem olduğu ve çözülmesi gerektiği üzerineydi. Burada mutlaka ve mutlaka https://t.co/5sQsXmeWFj incelenmeli. Ne yapılabilir sorusu twitter'ı aştı ve biz ken
@EJWagenmakers @BayesianSpecs Here's the opening sentence of "Estimating the reproducibility of psychological science" (https://t.co/rQe3Mcd6Bd 3rd in G.Sch. for "reproducibility"), IMHO W. OTOH 2nd on the list gets it right (https://t.co/F3Df5QyHJ0). ->
@XJRXZ Recomiendo ver el trabajo de Brian Nosek o Daniel Lakens sobre reproducibility y open science https://t.co/ka9IIFrsI7
@cabana_alvaro @MicrobiomDigest La crisis de replicación ya esta arriba nuestro y hay gente haciendo cosas al respecto, recomiendo Nosek, B. (2015) https://t.co/Oi9ylvcNLe Calculo q en Uruguay empezaremos a debatir esto en unos 30 años, cuando le aflojemos
@Oggie65 @mudbutt42 @jonnyenviro @sportsjunkie007 @StickProfessor "Replication effects were half the magnitude of original effects, representing substantial decline. Ninety-seven percent of original studies had statistically significant results. Thirty-six
@wazwords There is an interesting project attempting to recreate psychology experiments 'Estimating the reproducibility of psychological science' Just because you learnt it at Uni "It ain't necessarily so" https://t.co/Sstf0AfJKa
@dstaubli @Salvesalvi Here you go: https://t.co/9Qpo6dGtcT Experimental econ fares a bit better: https://t.co/sBoJgzLvNE But some might say that's just b/c it also tests a lot of obvious hypotheses😉
I want to cite this tweet in future revision memos :)
We know people think & behave differently. (Obv, right?) So, if effects are uniform, they don’t explain observed variation. That’s fine for individual tests. But if we think that’s true for *everything* then we haven’t answered the big questions. Soc I
RT @A_agadjanian: Given how hard it is to detect interaction effects (eg https://t.co/JFv2URNGTJ) & how often they don't replicate (eg only…
RT @A_agadjanian: Given how hard it is to detect interaction effects (eg https://t.co/JFv2URNGTJ) & how often they don't replicate (eg only…
RT @A_agadjanian: Given how hard it is to detect interaction effects (eg https://t.co/JFv2URNGTJ) & how often they don't replicate (eg only…
RT @A_agadjanian: Given how hard it is to detect interaction effects (eg https://t.co/JFv2URNGTJ) & how often they don't replicate (eg only…
RT @A_agadjanian: Given how hard it is to detect interaction effects (eg https://t.co/JFv2URNGTJ) & how often they don't replicate (eg only…
RT @A_agadjanian: Given how hard it is to detect interaction effects (eg https://t.co/JFv2URNGTJ) & how often they don't replicate (eg only…
I totally knew this. ^^
RT @A_agadjanian: Given how hard it is to detect interaction effects (eg https://t.co/JFv2URNGTJ) & how often they don't replicate (eg only…
RT @A_agadjanian: Given how hard it is to detect interaction effects (eg https://t.co/JFv2URNGTJ) & how often they don't replicate (eg only…
RT @A_agadjanian: Given how hard it is to detect interaction effects (eg https://t.co/JFv2URNGTJ) & how often they don't replicate (eg only…
RT @A_agadjanian: Given how hard it is to detect interaction effects (eg https://t.co/JFv2URNGTJ) & how often they don't replicate (eg only…
RT @chrishanretty: "Most effects are small" and "fuck nuance" appear in the epic cross-over hit, "tell me why you *need* that interaction"…
"Most effects are small" and "fuck nuance" appear in the epic cross-over hit, "tell me why you *need* that interaction"
RT @A_agadjanian: Given how hard it is to detect interaction effects (eg https://t.co/JFv2URNGTJ) & how often they don't replicate (eg only…
RT @A_agadjanian: Given how hard it is to detect interaction effects (eg https://t.co/JFv2URNGTJ) & how often they don't replicate (eg only…
Given how hard it is to detect interaction effects (eg https://t.co/JFv2URNGTJ) & how often they don't replicate (eg only 22% success here https://t.co/FYDRm9it03), I don't really get how "do effects vary by [moderator x]" remains one of the more commo
RT @BrianNosek: Facebook reminded me that a press release critiquing the Reproducibility Project: Psychology (https://t.co/kgCfKw6DjS) came…
RT @BrianNosek: Facebook reminded me that a press release critiquing the Reproducibility Project: Psychology (https://t.co/kgCfKw6DjS) came…
RT @BrianNosek: Facebook reminded me that a press release critiquing the Reproducibility Project: Psychology (https://t.co/kgCfKw6DjS) came…
RT @BrianNosek: Facebook reminded me that a press release critiquing the Reproducibility Project: Psychology (https://t.co/kgCfKw6DjS) came…
RT @BrianNosek: Facebook reminded me that a press release critiquing the Reproducibility Project: Psychology (https://t.co/kgCfKw6DjS) came…
RT @BrianNosek: Facebook reminded me that a press release critiquing the Reproducibility Project: Psychology (https://t.co/kgCfKw6DjS) came…
RT @BrianNosek: Facebook reminded me that a press release critiquing the Reproducibility Project: Psychology (https://t.co/kgCfKw6DjS) came…
Facebook reminded me that a press release critiquing the Reproducibility Project: Psychology (https://t.co/kgCfKw6DjS) came out 5 years ago today. This seems like an appropriate occasion to reflect on the critique and evidence that has accumulated since.
@pechotbourdon @georgevanhal @ionicasmeets Precies - dat is natuurlijk überhaupt de functie van papers, om te overtuigen dat iets wel of niet zo is, dat is het dramatische gevolg van de huidige publicatiecultuur. Hier wat onderzoek naar de reproductiecrisi
Estimating the reproducibility of psychological science https://t.co/0HgxdenQUj
Estimating the reproducibility of psychological science https://t.co/weWkRhJhXd
RT @T_TEST__: 心理学(研究)に興味があるならばOpen Science Collaboration (2015) Scienceは知っといたほうがいいね。 無作為に選んだ心理学研究の論文の追試をしたら統計結果を再現できたのが4割弱しかなかったというやつ。 Est…
心理学(研究)に興味があるならばOpen Science Collaboration (2015) Scienceは知っといたほうがいいね。 無作為に選んだ心理学研究の論文の追試をしたら統計結果を再現できたのが4割弱しかなかったというやつ。 Estimating the reproducibility of psychological science https://t.co/EDO7ouMUeg
2/2 Let's also remind ourselves, that a big part of research in the field of clinical psychology is rubbish. https://t.co/Z9jboK2Fyx
@RodanHassan @SirJoeberg @DuNyaDuFria Statsvetenskap är också en av de vetenskaper som inte gör replikationer, så en massa i läroböckerna är falskt. Denna artikel är intressant. Man upprepade 100 psykologiska experiment, hälften gick att replikera! https
Pourquoi répliquer les réplications ? Vous savez sûrement que les précédentes réplications (impliquant la collab de plusieurs labos et des gros échantillons) ont rarement réussi à répliquer des résultats parfois très connus (entre 20% et 40% de succès). ht
@naval @cvaldary "mostly fake" seems to me a mischaracterization, but there might indeed be a reproducibility problem, as there is in other disciplines: https://t.co/w0CCqBlGaw https://t.co/ltQZRgeKk0 https://t.co/iQelveRH8K
For this class, I was inspired by @OrbenAmy 's syllabus https://t.co/LpjNDFWfu1 that uses several platforms to talk about this topic. Here is how the class was structured: First, the students read the 2015 @OSFramework paper: https://t.co/9QS4ZG25Q8 3/
Estimating the reproducibility of psychological science https://t.co/F3sAlcW84s
@iamproblematic2 @topspin7777 @RandPaul @realDonaldTrump That’s for psychology, champ. https://t.co/IGxHrVaTjD
Starting in a few minutes at 4 pm Brasília, @ReproducibiliT Brasil #5 discusses the Reproducibility Project: Psychology (https://t.co/E38igDNRkR). Join us live at https://t.co/5ydbCpkxxj (in Portuguese)! https://t.co/jtE3bszEXS
論文の知見がすべて真実ではないということを実感するためにこの論文も読んでもらいたいが,学部生さんが読むにはしんどいかな...💦 自分が最初に紹介しようかな Estimating the reproducibility of psychological science https://t.co/lS0cy6K9zX
RT @DeBunKerEtoiles: Reproductibility project: Des chercheurs ont vérifié la reproductibilité de 100 études de psychologie publiées dans de…
RT @DeBunKerEtoiles: Reproductibility project: Des chercheurs ont vérifié la reproductibilité de 100 études de psychologie publiées dans de…
RT @DeBunKerEtoiles: Reproductibility project: Des chercheurs ont vérifié la reproductibilité de 100 études de psychologie publiées dans de…
RT @DeBunKerEtoiles: Reproductibility project: Des chercheurs ont vérifié la reproductibilité de 100 études de psychologie publiées dans de…
RT @DeBunKerEtoiles: Reproductibility project: Des chercheurs ont vérifié la reproductibilité de 100 études de psychologie publiées dans de…
RT @DeBunKerEtoiles: Reproductibility project: Des chercheurs ont vérifié la reproductibilité de 100 études de psychologie publiées dans de…
RT @DeBunKerEtoiles: Reproductibility project: Des chercheurs ont vérifié la reproductibilité de 100 études de psychologie publiées dans de…
RT @DeBunKerEtoiles: Reproductibility project: Des chercheurs ont vérifié la reproductibilité de 100 études de psychologie publiées dans de…
RT @DeBunKerEtoiles: Reproductibility project: Des chercheurs ont vérifié la reproductibilité de 100 études de psychologie publiées dans de…
RT @DeBunKerEtoiles: Reproductibility project: Des chercheurs ont vérifié la reproductibilité de 100 études de psychologie publiées dans de…
RT @DeBunKerEtoiles: Reproductibility project: Des chercheurs ont vérifié la reproductibilité de 100 études de psychologie publiées dans de…
RT @DeBunKerEtoiles: Reproductibility project: Des chercheurs ont vérifié la reproductibilité de 100 études de psychologie publiées dans de…
RT @DeBunKerEtoiles: Reproductibility project: Des chercheurs ont vérifié la reproductibilité de 100 études de psychologie publiées dans de…
Reproductibility project: Des chercheurs ont vérifié la reproductibilité de 100 études de psychologie publiées dans des revues à comité de lecture. L’effet allégué a pu être reproduit dans 39% des cas. N’oubliez pas qu’une étude seule n’est pas sûre à 100%
RT @EmoFree: دراسة قامت بإعادة تطبيق ١٠٠ دراسة حديثة في علم النفس والتي نشرت في أفضل المجلات العلمية المحكمة. وبينت الدراسة أن ٣٩ دراسة…
دراسة قامت بإعادة تطبيق ١٠٠ دراسة حديثة في علم النفس والتي نشرت في أفضل المجلات العلمية المحكمة. وبينت الدراسة أن ٣٩ دراسة فقط أعطت نفس نتائج الدراسة الأساسية. https://t.co/KOznK0Xdzx
@WvNoort Uit wetenschappelijk onderzoek blijkt dat sociaal-psychologisch onderzoek meestal de toets der kritiek niet kan doorstaan. https://t.co/1QtMYgrxnl
@Will_Cashmore @nopeOZ @Sawdust43336012 @JohnB250379 @prageru Also, "science" today is so unreliable, the conclusions are next to useless. A product of philosophers denying the scientific method on mind matters. If the correlations are given, the data can
@CptAntiRacist @Spacewitch024 @griffonatrix @LoFiRepublican "Correlational tests suggest that replication success was better predicted by the strength of original evidence than by characteristics of the original and replication teams." Gee, who could've gu
You should not simply rely on p-values, Hunter. Holding onto p<0.05 does not do much if you don't even know the studies effect size. This is why statisticians are losing hope on p-values since studies using the 0.05 threshold can barely replicate https
@SlowMoneyGreen @shesmegg82 @Dcoronata @GeraldKutney @_LoveMTB_ I didn’t make the rules...repeatability is a key aspect of good science no? You keep up the ad hominem, I’ll keep making you look like the dumb one... https://t.co/uG7i38vrbk
RT @BIOsloTea: For our ReproducibiliTea meeting this Thursday, we will discuss the reproducibility of psychological science. https://t.co/…
For our ReproducibiliTea meeting this Thursday, we will discuss the reproducibility of psychological science. https://t.co/e2rSj76HmD Want to join? Send us a DM for the Zoom link😀
Replication effects were half the magnitude of original effects, representing a substantial decline https://t.co/bDw1a4DPkA
RT @RyoMasahiro: How reproducible p-values are? They are totally not!! I was gasped when I saw this figure, comparing p-values reported in…
RT @RyoMasahiro: How reproducible p-values are? They are totally not!! I was gasped when I saw this figure, comparing p-values reported in…
RT @RyoMasahiro: How reproducible p-values are? They are totally not!! I was gasped when I saw this figure, comparing p-values reported in…