Synopsis of Social media discussions
The discussions include mentions of redefining key statistical concepts like 'confidence intervals' as 'compatibility intervals' and share insights on the cognitive implications of statistical interpretations. Phrases such as 'must consider' and references to the publication as a necessary read highlight the urgency and seriousness with which many posters view the potential shift in statistical methodology.
Agreement
Moderate agreementMost discussions reflect a supportive stance towards the publication's recommended changes in statistical tools.
Interest
High level of interestThe posts reveal high interest, with many users engaging in detailed discussions about the implications of the methodologies proposed in the article.
Engagement
High engagementMany contributors offer thoughtful analyses or critiques, demonstrating a deep level of engagement with the article's content.
Impact
High level of impactNumerous commenters believe that the publication will significantly influence statistical practices and education.
Social Mentions
YouTube
2 Videos
2 Posts
149 Posts
Blogs
2 Articles
2 Posts
Metrics
Video Views
2,210
Total Likes
378
Extended Reach
880,212
Social Features
157
Timeline: Posts about article
Top Social Media Posts
さくら
95％信頼区間も話題ですね。
P値や信頼区間の解釈を変えた方が良い、という趣旨の論文を近年多く見かけます。
旧式の（？）良さを残しつつモデルとデータの適合度compatibilityという概念もしっかり理解したいですね（私も勉強中です）。
https://t.co/7TgFhk6zMD
https://t.co/VVNPYhsWfp
view full post
Sir Panda (Zad Rafi)
@DataSciFact @JohnDCook It’s true, and there are so many other misconceptions regarding what hypotheses you can generate p values for and many misconceptions of assumptions. Here’s a definition we give in our paper: https://t.co/uUDid9Iv4T https://t.co/i9FpvzV78w
view full post
Sander Greenland
@david_colquhoun @vamrhein @dailyzad @learnfromerror David, would you please be clear that our call is to retire deceptive terms like "significance" and "confidence" to describe P-values and intervals based on P-value functions (which show the compatibility of effect sizes with data), NOT to retire P-values https://t.co/Tnq4bZSoAQ
view full post
黒木玄 Gen Kuroki
#統計
https://t.co/MxoRcISI2B
Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surprise
Z. Rafi, S. Greenland
BMC medical research methodology, 2020
の音声概要を作り直しました。v2.0
倍程度長くなっています。 https://t.co/edVLDBzMvA https://t.co/xgejVsa6tc
view full post
Sir Panda (Zad Rafi)
Glad to see this finally up, it’s open access so anyone can read it and has been in the works for nearly two years, partly due to the amount of helpful feedback and comments we’ve received. 1/n
link: https://t.co/h21MqyylSO
enhanced PDF: https://t.co/1CTzHHhBgY https://t.co/Mau8GBd1jw
view full post
Posts referencing the article
Rethinking Statistical Analysis: Compatibility Over Confidence
Researchers often misinterpret statistical outputs, leading to a demand for new methods replacing traditional P-values. This video reviews cognitive aspects influencing such misinterpretations and presents an argument for reinterpreting P-values as measures of compatibility.
July 17, 2023
2,072 views
Redefining P-Values: Understanding Compatibility in Statistics
This video presents a guided example addressing common misinterpretations of P-values. Rooted in recent research, it explores the benefits of viewing P-values as measures of compatibility, enhancing researchers' understanding of data analysis.
June 4, 2023
138 views
-
RT @genkuroki: #統計 NotebookLMで作った以下の論文の動画解説 https://t.co/MxoRcISI2B Semantic and cognitive tools to aid statistical science: replace confi…
view full postSeptember 11, 2025
3
-
備忘録＋α読書感想
@krdw2 (Twitter)RT @genkuroki: #統計 NotebookLMで作った以下の論文の動画解説 https://t.co/MxoRcISI2B Semantic and cognitive tools to aid statistical science: replace confi…
view full postSeptember 11, 2025
3
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @Lester_Domes: @david_colquhoun @vamrhein @dailyzad @learnfromerror David, would you please be clear that our call is to retire deceptiv…
view full postSeptember 6, 2025
9
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @genkuroki: #統計 NotebookLMで作った以下の論文の動画解説 https://t.co/MxoRcISI2B Semantic and cognitive tools to aid statistical science: replace confi…
view full postAugust 27, 2025
3
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)#統計 NotebookLMで作った以下の論文の動画解説 https://t.co/MxoRcISI2B Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surprise Z. Rafi, S. Greenland BMC medical research methodology, 2020 https://t.co/Ln5X0NzCwQ https://t.co/F1yhNNbvLk
view full postAugust 26, 2025
6
3
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @genkuroki: #統計 https://t.co/MxoRcISI2B Semantic and cognitive tools to aid statistical science: replace confidence and significance by…
view full postAugust 13, 2025
6
-
清水 団 Dan Shimizu
@dannchu (Twitter)RT @genkuroki: #統計 https://t.co/MxoRcISI2B Semantic and cognitive tools to aid statistical science: replace confidence and significance by…
view full postJuly 23, 2025
6
-
akakit
@akakit (Twitter)RT @genkuroki: #統計 https://t.co/MxoRcISI2B Semantic and cognitive tools to aid statistical science: replace confidence and significance by…
view full postJuly 23, 2025
6
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @genkuroki: #統計 https://t.co/MxoRcISI2B Semantic and cognitive tools to aid statistical science: replace confidence and significance by…
view full postJuly 16, 2025
6
-
磯崎 ゆい
@hirohito_ (Twitter)RT @genkuroki: #統計 https://t.co/MxoRcISI2B Semantic and cognitive tools to aid statistical science: replace confidence and significance by…
view full postJuly 9, 2025
6
-
Ilfan Oh, PhD (오일환 呉 日煥 オ・イルファン)
@Ilfan_Oh (Twitter)RT @genkuroki: #統計 https://t.co/MxoRcISI2B Semantic and cognitive tools to aid statistical science: replace confidence and significance by…
view full postJuly 1, 2025
6
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @Lester_Domes: @david_colquhoun @vamrhein @dailyzad @learnfromerror David, would you please be clear that our call is to retire deceptiv…
view full postJuly 1, 2025
9
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)#統計 https://t.co/MxoRcISI2B Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surprise Z. Rafi, S. Greenland BMC medical research methodology, 2020 の音声概要を作り直しました。v2.0 倍程度長くなっています。 https://t.co/edVLDBzMvA https://t.co/xgejVsa6tc
view full postJune 30, 2025
12
6
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @genkuroki: #統計 過信の原因になる「信頼」と「有意性」の概念を「相性の良さ」(compatibility)に置き換える提案 https://t.co/1qdgTcOKEK Semantic and cognitive tools to aid statis…
view full postJune 21, 2025
2
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @genkuroki: #統計 過信の原因になる「信頼」と「有意性」の概念を「相性の良さ」(compatibility)に置き換える提案 https://t.co/1qdgTcOKEK Semantic and cognitive tools to aid statis…
view full postMay 11, 2025
2
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @genkuroki: #統計 添付画像はRafi-Greenland 2020 https://t.co/1qdgTcPiui より。 P値=0.9999とP値=0.90の比は0.9999/0.90=1.111で1に近く、P値=0.10とP値=0.0001の比は0.1…
view full postMay 7, 2025
1
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @genkuroki: #統計 Rafi-Greenland 2020 https://t.co/1qdgTcOKEK のsurprisal value = S値について。S値は S値 = -log₂ P値 (単位はビット) と定義されます。P値の違いの大きさはS値の…
view full postMay 7, 2025
1
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @genkuroki: #統計 巷では統計学通を騙る半可通達が P値ではなく信頼区間を使える(ニヤ) と言っているようだが、既出の論文 https://t.co/1qdgTcPiui Rafi-Greenland 2020 は信頼区間も悪しき二分法の原因になって…
view full postMay 7, 2025
2
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)#統計 過信の原因になる「信頼」と「有意性」の概念を「相性の良さ」(compatibility)に置き換える提案 https://t.co/1qdgTcOKEK Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surprise Zad Rafi & Sander Greenland 2020 https://t.co/RiBTl4074a https://t.co/tyxmIWKRiM
view full postMay 6, 2025
2
2
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @genkuroki: #統計 添付画像はRafi-Greenland 2020 https://t.co/1qdgTcPiui より。 P値=0.9999とP値=0.90の差と、P値=0.10とP値=0.0001の差は、どちらも0.0999なのですが、P値の差でP値の…
view full postMay 6, 2025
1
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @genkuroki: #統計 Rafi-Greenland 2020 https://t.co/1qdgTcOKEK の音声概要 https://t.co/7tGGHWuEWe (9分) https://t.co/8fMSN2X1TH
view full postMay 6, 2025
2
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)#統計 添付画像はRafi-Greenland 2020 https://t.co/1qdgTcPiui より。 P値=0.9999とP値=0.90の比は0.9999/0.90=1.111で1に近く、P値=0.10とP値=0.0001の比は0.10/0.0001=1000で非常に大きい。 P値=0.9999とP値=0.90の違いは小さく、P値=0.10とP値=0.0001の違いは非常に大きい。続く https://t.co/LlwgVTZiSo
view full postMay 6, 2025
4
1
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)#統計 添付画像はRafi-Greenland 2020 https://t.co/1qdgTcPiui より。 P値=0.9999とP値=0.90の差と、P値=0.10とP値=0.0001の差は、どちらも0.0999なのですが、P値の差でP値の違いの大きさを測るのは誤りになります。 P値の違いの大きさは、差ではなく、比で測る必要があります。続く https://t.co/UNhNCRAjtb
view full postMay 6, 2025
4
1
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)#統計 Rafi-Greenland 2020 https://t.co/1qdgTcOKEK のsurprisal value = S値について。S値は S値 = -log₂ P値 (単位はビット) と定義されます。P値の違いの大きさはS値の違いの大きさで見るべきです。P値のままだと一般人は違いの程度を誤解し易い。続く https://t.co/ivl5fGRsN3 https://t.co/xgejVs9yDE
view full postMay 6, 2025
3
1
-
TT
@reptrt (Twitter)RT @genkuroki: #統計 巷では統計学通を騙る半可通達が P値ではなく信頼区間を使える(ニヤ) と言っているようだが、既出の論文 https://t.co/1qdgTcPiui Rafi-Greenland 2020 は信頼区間も悪しき二分法の原因になって…
view full postMay 5, 2025
2
-
TT
@reptrt (Twitter)RT @genkuroki: #統計 Rafi-Greenland 2020 https://t.co/1qdgTcOKEK の音声概要 https://t.co/7tGGHWuEWe (9分) https://t.co/8fMSN2X1TH
view full postMay 5, 2025
2
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)#統計 巷では統計学通を騙る半可通達が P値ではなく信頼区間を使える(ニヤ) と言っているようだが、既出の論文 https://t.co/1qdgTcPiui Rafi-Greenland 2020 は信頼区間も悪しき二分法の原因になっていることが分かるように書かれています。 半可通達は多分この手の論文を読もうとしていない。 https://t.co/xgejVsa6tc
view full postMay 5, 2025
3
2
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)#統計 Rafi-Greenland 2020 https://t.co/1qdgTcOKEK の音声概要 https://t.co/7tGGHWuEWe (9分) https://t.co/8fMSN2X1TH
view full postMay 5, 2025
3
2
-
hr_wksh
@hr_wksh (Twitter)RT @ai_biostat: 「有意性」「信頼区間」という用語の代わりに「互換性（compatibility）」「互換性区間（compatibility interval）」という表現を提案しています。音声リンクはこちら
view full postMay 4, 2025
4
-
mocchi
@mocchi_ym (Twitter)RT @ai_biostat: 「有意性」「信頼区間」という用語の代わりに「互換性（compatibility）」「互換性区間（compatibility interval）」という表現を提案しています。音声リンクはこちら
view full postMay 4, 2025
4
-
ローレライ
@usagii124 (Twitter)RT @ai_biostat: 「有意性」「信頼区間」という用語の代わりに「互換性（compatibility）」「互換性区間（compatibility interval）」という表現を提案しています。音声リンクはこちら
view full postMay 4, 2025
4
-
わたヤク@AI臨床研究
@ai_biostat (Twitter)「有意性」「信頼区間」という用語の代わりに「互換性（compatibility）」「互換性区間（compatibility interval）」という表現を提案しています。音声リンクはこちら
view full postMay 4, 2025
9
4
-
Katsushi Kagaya @katzkagaya.bsky.social
@katzkagaya (Twitter)RT @genkuroki: #統計 論文へのリンク https://t.co/1qdgTcPiui Semantic and cognitive tools to aid statistical science: replace confidence and signifi…
view full postApril 7, 2025
1
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)#統計 論文へのリンク https://t.co/1qdgTcPiui Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surprise Zad Rafi & Sander Greenland 2020 大体においてP値に関する正しい考え方についてはGreenlandさん達の論文がお勧め。 https://t.co/HWsRJYZF4N
view full postApril 7, 2025
2
1
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @Lester_Domes: @david_colquhoun @vamrhein @dailyzad @learnfromerror David, would you please be clear that our call is to retire deceptiv…
view full postMarch 27, 2025
9
-
Alex Leaf
@AlexJLeaf (Twitter)@HRH_Ted Based on your comment, I'm guessing you don't know what a confidence interval is or what it represents. Here's a good summary review to read, if interested. https://t.co/WcpIDBxscH
view full postMarch 1, 2025
3
-
uncorrelated
@uncorrelated (Twitter)Rafi and Greenland (2020) https://t.co/jhGDqoW3up で、“confidence” refers to the procedure behavior, not the reported interval, seems to be lost on most researchersと、CI percentile levelは、報告された区間ではなく、区間をつくる手続き（procedure）に対してだと説明していたり。 https://t.co/HSr5HUIRb3
view full postJanuary 10, 2025
6
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @Lester_Domes: @david_colquhoun @vamrhein @dailyzad @learnfromerror David, would you please be clear that our call is to retire deceptiv…
view full postDecember 22, 2024
9
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @Lester_Domes: @david_colquhoun @vamrhein @dailyzad @learnfromerror David, would you please be clear that our call is to retire deceptiv…
view full postNovember 29, 2024
9
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @Lester_Domes: @david_colquhoun @vamrhein @dailyzad @learnfromerror David, would you please be clear that our call is to retire deceptiv…
view full postOctober 23, 2024
9
-
Daniel
@strengejacke (Twitter)@mzloteanu "We can now define an observed P-value p as the probability of the test statistic being at least as extreme as observed if the hypothesis H targeted for testing and every assumption used to compute the P-value (...) were correct" (10.1186/s12874-020-01105-9)
view full postSeptember 17, 2024
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @Lester_Domes: @learnfromerror @Abraham_RMI @trumanfrancis @vamrhein @lakens @f2harrell @stephensenn @goodmanmetrics @statsepi Sorry: pu…
view full postAugust 15, 2024
2
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @Lester_Domes: @david_colquhoun @vamrhein @dailyzad @learnfromerror David, would you please be clear that our call is to retire deceptiv…
view full postAugust 15, 2024
9
-
ニャンコ先生
@wakadaishou (Twitter)その後にGreenland氏の論文がいくつか紹介されていたので読んでみました。 Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surprise Rafi and Greenland BMC Medical Research Methodology (2020) 20:244
view full postApril 16, 2024
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)RT @strengejacke: @Brier_Gallihugh @camjpatrick @prisonrodeo @carlislerainey @bmwiernik https://t.co/luvFvZbSwV https://t.co/7x7Xk4Xfdg h…
view full postMarch 8, 2024
1
-
Daniel
@strengejacke (Twitter)RT @dailyzad: @f2harrell @vandy_biostat @EdgeforScholars Further discussion here: https://t.co/DepyiqagSt https://t.co/E5gzPha7Hh
view full postMarch 8, 2024
1
-
Daniel
@strengejacke (Twitter)@Brier_Gallihugh @camjpatrick @prisonrodeo @carlislerainey @bmwiernik https://t.co/luvFvZbSwV https://t.co/7x7Xk4Xfdg https://t.co/WdfweoVZnx You find some of them on Twitter, too (@dailyzad or @Lester_Domes , don't know about the others)
view full postMarch 8, 2024
2
1
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @Lester_Domes: @jonathanasterne @mendel_random @Caketin_Wade @ruthkipping Define P-values first and generally, for any parameter. Then:…
view full postFebruary 11, 2024
4
-
Smit Giri MD MHS
@smith__giri (Twitter)RT @PavlosMsaouel:
view full postJanuary 12, 2024
2
-
Isabella R. Ghement
@IsabellaGhement (Twitter)RT @PavlosMsaouel:
view full postJanuary 11, 2024
2
-
Pavlos Msaouel
@PavlosMsaouel (Twitter)January 11, 2024
7
2
-
Sander Greenland
@Lester_Domes (Twitter)@drjgauthier @f2harrell @MaartenvSmeden @GSCollins @Richard_D_Riley @StatModeling 2>Warranted uncertainty is lower when selection for observation and treatment are random rather than merely unknown. Thus the method is founded on the same best-case scenario as in almost all stat methods and risks creating false certainty (overconfidence) https://t.co/Tnq4bZSoAQ
view full postDecember 17, 2023
6
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)RT @genkuroki: #統計 Greenlandさん達が提案している改革の1つは、P値を データの数値 統計モデル パラメータ値の設定に関する仮説 の相性の良さ(compatibility)の指標の1つだと説明することです。 これなら相性を見る対象達の全てを気楽に…
view full postNovember 19, 2023
1
-
黒木玄 Gen Kuroki
@genkuroki (Twitter)#統計 Greenlandさん達が提案している改革の1つは、P値を データの数値 統計モデル パラメータ値の設定に関する仮説 の相性の良さ(compatibility)の指標の1つだと説明することです。 これなら相性を見る対象達の全てを気楽に疑えます。 https://t.co/1qdgTcPiui Zad Rafi & Sander Greenland 2020
view full postNovember 19, 2023
7
1
-
Zach Hopkins, MD
@ZachHopkinsMD (Twitter)RT @doc_BLocke: @ZachHopkinsMD Framing I like better: P val summarize how surprising a result would be *if there's nothing going on* https:…
view full postSeptember 8, 2023
1
-
Brian Locke, MD MSCI
@doc_BLocke (Twitter)@ZachHopkinsMD Framing I like better: P val summarize how surprising a result would be *if there's nothing going on* https://t.co/s778sdIzYN If too surprising to be nothing, must eval if model/exp design misspec (confounding, etc) or evidence - but it's only evidence by process of elimination
view full postSeptember 8, 2023
2
1
-
Sander Greenland
@Lester_Domes (Twitter)@analisereal @lakens @CourtneyP_PhD @BradSchoenfeld @ChrisSchneiRUB 5. Addenda to 1-4 above: #2 should say "S-values make clear why p=.01 and p=0.05 are far apart yet p=0.95 AND p=0.99 are not." Rafi & Greenland is free at https://t.co/Tnq4bZSWqo Those into theory may also find of interest https://t.co/MXTzWpjnqY and https://t.co/4Z2EzDgwIn
view full postJuly 27, 2023
3
2
-
Jake Elder
@Jake_Elder52 (Twitter)@Brier_Gallihugh https://t.co/eWNQf03vBX
view full postJune 19, 2023
1
-
Pavlos Msaouel
@PavlosMsaouel (Twitter)@BrianCHolt It has been explained ad nauseam, including by the person who invented CIs (eg
view full postMay 28, 2023
4
-
Oli Pescott
@sacrevert (Twitter)@dsquintana E.g. https://t.co/DWsKy3pdHR
view full postMarch 31, 2023
-
Cyrus Samii
@cdsamii (Twitter)@simonhhess @Andrew___Baker @DogmaticPrior See the P-value function graph in Rafi and Greenland: https://t.co/slwpV9s0ct
view full postMarch 14, 2023
1
-
Gregory Lopez
@GLopezPharmD (Twitter)@AaronBergman18 I like the approach of computing for multiple hypotheses of interest and/or interpret CIs as compatibility intervals. The latter's particularly useful for interpreting existing research: https://t.co/pzM510mnSK
view full postFebruary 18, 2023
1
-
Henry Bemis
@henrybemis13 (Twitter)@gehsbarg @hempelian @ajordannafa @_____Lightning @statwonk P values are fine. Here’s one way of thinking about them https://t.co/lFFhE7X4Ld
view full postFebruary 1, 2023
-
epibot
@epi_twit (Twitter)RT @connorbgrady: https://t.co/QrNLwkFKXm Have you used recommendations from Rafi and Greenland in published work? →"compatibility" over "…
view full postDecember 28, 2022
2
-
Rstats
@rstatstweet (Twitter)RT @connorbgrady: https://t.co/QrNLwkFKXm Have you used recommendations from Rafi and Greenland in published work? →"compatibility" over "…
view full postDecember 28, 2022
2
-
Connor Grady
@connorbgrady (Twitter)https://t.co/QrNLwkFKXm Have you used recommendations from Rafi and Greenland in published work? →"compatibility" over "significance" →"compatibility intervals" over "confidence intervals" →S-values →P-values for alternative hypotheses #RStats #Epitwitter #MedTwitter #Stats
view full postDecember 28, 2022
3
2
-
Jane Shevtsov
@jaiagreen (Twitter)@TAH_Sci @PaulHenryTremb1 @apsmunro The really informative way to use a CI is to just interpret it as what values of the parameter measured (relative risk, in this case) are compatible with the data. The hypothesis is irrelevant. https://t.co/1mibD1UvTy
view full postDecember 1, 2022
1
-
Mark Annuncio
@MarkAnnuncio (Twitter)@IsabellaGhement @mzelst @Ryan_ETyler @EikoFried @bolkerb I assume you are familiar with Greenland, Rafi, and co work? If not, I would say it addresses what you are saying https://t.co/jZUvnB39S6 https://t.co/ObT2qhiSXE
view full postNovember 5, 2022
1
-
Dylan Armbruster
@dylanarmbruste3 (Twitter)@ScepticalDoctor @MyNutritionSci https://t.co/k5Tj7ibkMV This paper from Rafi and Greenland provide an exampe.example. I'm sure I could link many others as well.
view full postOctober 18, 2022
-
みのゆ
@minoyu (Twitter)某所の信頼区間論争、CIが0を跨いだから急に意味が変わるわけではないというのはその通りだと思う。特に以下のRafiとGreenlandのいうcompatibility解釈であれば。 こういうの見てるとみんなベイズにしようぜという気持ちになってくる。 https://t.co/gCYmgbKXAf
view full postOctober 15, 2022
2
-
Pavlos Msaouel
@PavlosMsaouel (Twitter)@arnabguonc @Uromigos @DrChoueiri @myESMO @brian_rini @arihakimi @johnleppert @kcCURE @BradLeibovichMD @MVossMD @DrDavidNanus @OAlhalabiMD @DrScottHaake @katy_beckermann Not exactly, remember that (frequentist) CIs are functions of P-values. Significance is a decision-theoretic concept different than either CIs or p-values. To help unpack all these start
view full postSeptember 28, 2022
1
-
Dana Zeid
@zeid_dana (Twitter)RT @bmwiernik: @ABindoff @JenWenzel1 @NeuroBill This paper has great discussion of more useful and accessible ways to describe statistical…
view full postSeptember 26, 2022
1
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)RT @Lester_Domes: @avicenna @conjugateprior @vamrhein Among them are warnings against misinterpretations in Pearson 1906 p. 183, Boring 191…
view full postSeptember 26, 2022
1
-
Brenton Wiernik
@bmwiernik (Twitter)@ABindoff @JenWenzel1 @NeuroBill This paper has great discussion of more useful and accessible ways to describe statistical results https://t.co/5H9SgdOzmN
view full postSeptember 16, 2022
3
1
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)@orpanag @omaclaren @richarddmorey @jd_wilko @statsepi We talk about this issue on pg 11 here https://t.co/XkbibhpW8d https://t.co/sAMvS7xVFL
view full postSeptember 15, 2022
1
-
pricing.o
@o_pricing (Twitter)RT @Sakura_Med_DSci: 95％信頼区間も話題ですね。 P値や信頼区間の解釈を変えた方が良い、という趣旨の論文を近年多く見かけます。 旧式の（？）良さを残しつつモデルとデータの適合度compatibilityという概念もしっかり理解したいですね（私も勉強中です…
view full postAugust 28, 2022
10
-
murapo
@murapo1 (Twitter)RT @Sakura_Med_DSci: 95％信頼区間も話題ですね。 P値や信頼区間の解釈を変えた方が良い、という趣旨の論文を近年多く見かけます。 旧式の（？）良さを残しつつモデルとデータの適合度compatibilityという概念もしっかり理解したいですね（私も勉強中です…
view full postAugust 28, 2022
10
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)RT @Sakura_Med_DSci: 95％信頼区間も話題ですね。 P値や信頼区間の解釈を変えた方が良い、という趣旨の論文を近年多く見かけます。 旧式の（？）良さを残しつつモデルとデータの適合度compatibilityという概念もしっかり理解したいですね（私も勉強中です…
view full postAugust 28, 2022
10
-
hajk__
@hajk__ (Twitter)RT @Sakura_Med_DSci: 95％信頼区間も話題ですね。 P値や信頼区間の解釈を変えた方が良い、という趣旨の論文を近年多く見かけます。 旧式の（？）良さを残しつつモデルとデータの適合度compatibilityという概念もしっかり理解したいですね（私も勉強中です…
view full postAugust 28, 2022
10
-
ysaito
@ysaito8015 (Twitter)RT @Sakura_Med_DSci: 95％信頼区間も話題ですね。 P値や信頼区間の解釈を変えた方が良い、という趣旨の論文を近年多く見かけます。 旧式の（？）良さを残しつつモデルとデータの適合度compatibilityという概念もしっかり理解したいですね（私も勉強中です…
view full postAugust 27, 2022
10
-
kokako
@kokako65541706 (Twitter)RT @Sakura_Med_DSci: 95％信頼区間も話題ですね。 P値や信頼区間の解釈を変えた方が良い、という趣旨の論文を近年多く見かけます。 旧式の（？）良さを残しつつモデルとデータの適合度compatibilityという概念もしっかり理解したいですね（私も勉強中です…
view full postAugust 27, 2022
10
-
Kazuhiro Tokuda （徳田和宏）
@tkd_k888 (Twitter)RT @Sakura_Med_DSci: 95％信頼区間も話題ですね。 P値や信頼区間の解釈を変えた方が良い、という趣旨の論文を近年多く見かけます。 旧式の（？）良さを残しつつモデルとデータの適合度compatibilityという概念もしっかり理解したいですね（私も勉強中です…
view full postAugust 27, 2022
10
-
さくら
@Sakura_Med_DSci (Twitter)RT @Sakura_Med_DSci: 95％信頼区間も話題ですね。 P値や信頼区間の解釈を変えた方が良い、という趣旨の論文を近年多く見かけます。 旧式の（？）良さを残しつつモデルとデータの適合度compatibilityという概念もしっかり理解したいですね（私も勉強中です…
view full postAugust 27, 2022
10
-
hr_wksh
@hr_wksh (Twitter)RT @Sakura_Med_DSci: 95％信頼区間も話題ですね。 P値や信頼区間の解釈を変えた方が良い、という趣旨の論文を近年多く見かけます。 旧式の（？）良さを残しつつモデルとデータの適合度compatibilityという概念もしっかり理解したいですね（私も勉強中です…
view full postAugust 27, 2022
10
-
まる
@standardtherap (Twitter)RT @Sakura_Med_DSci: 95％信頼区間も話題ですね。 P値や信頼区間の解釈を変えた方が良い、という趣旨の論文を近年多く見かけます。 旧式の（？）良さを残しつつモデルとデータの適合度compatibilityという概念もしっかり理解したいですね（私も勉強中です…
view full postAugust 27, 2022
10
-
佐藤わんこ@理系っぽい文系
@satowanko1 (Twitter)RT @Sakura_Med_DSci: 95％信頼区間も話題ですね。 P値や信頼区間の解釈を変えた方が良い、という趣旨の論文を近年多く見かけます。 旧式の（？）良さを残しつつモデルとデータの適合度compatibilityという概念もしっかり理解したいですね（私も勉強中です…
view full postAugust 27, 2022
10
-
さくら
@Sakura_Med_DSci (Twitter)95％信頼区間も話題ですね。 P値や信頼区間の解釈を変えた方が良い、という趣旨の論文を近年多く見かけます。 旧式の（？）良さを残しつつモデルとデータの適合度compatibilityという概念もしっかり理解したいですね（私も勉強中です）。 https://t.co/7TgFhk6zMD https://t.co/VVNPYhsWfp
view full postAugust 27, 2022
80
10
-
Abraham E. Gracia R.
@Abraham_RMI (Twitter)RT @vamrhein: @marcora @rlmcelreath Sorry, I can only reply by suggesting some literature. Compatibility is an old concept (e.g., Box 1980)…
view full postMay 20, 2022
2
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)RT @vamrhein: @marcora @rlmcelreath Sorry, I can only reply by suggesting some literature. Compatibility is an old concept (e.g., Box 1980)…
view full postMay 20, 2022
2
-
Valentin Amrhein
@vamrhein (Twitter)@marcora @rlmcelreath Sorry, I can only reply by suggesting some literature. Compatibility is an old concept (e.g., Box 1980), also called consonance (Folks 1981) and consistency (Cox 1977). Some references from our own work: https://t.co/56aYdYbA7D https://t.co/DEyXF8KL8h https://t.co/VUeUnitrZb
view full postMay 19, 2022
5
2
-
William Spagnola
@williamspagnola (Twitter)RT @com_ecology: @GodsoeWilliam The example really comes from other literature; so do it
view full postMay 14, 2022
2
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)RT @com_ecology: @GodsoeWilliam The example really comes from other literature; so do it
view full postMay 9, 2022
2
-
@comecology.bsky.social
@com_ecology (Twitter)@GodsoeWilliam The example really comes from other literature; so do it
view full postMay 9, 2022
5
2
-
Carlos E Lourenco
@caerib (Twitter)RT @dailyzad: @Lester_Domes And another example of the distinction. It's a much weaker condition/statement, so the point estimate being the…
view full postMay 8, 2022
1
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)@Lester_Domes And another example of the distinction. It's a much weaker condition/statement, so the point estimate being the most consistent with the data isn't that strong of a statement and is not that impressive, and rarely is it the truth https://t.co/DepyiqrRK1 https://t.co/doSL7dC7Qh
view full postMay 8, 2022
9
1
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)RT @williamspagnola: @tcarpenter216 @Lester_Domes and @dailyzad wrote a paper on semantic tools for understanding p-values given that the t…
view full postApril 9, 2022
1
-
William Spagnola
@williamspagnola (Twitter)@tcarpenter216 @Lester_Domes and @dailyzad wrote a paper on semantic tools for understanding p-values given that the technical definition just isn’t very intuitive https://t.co/metj3SXyAf
view full postApril 5, 2022
3
1
-
Alex Kale
@AlexKale17 (Twitter)RT @JessicaHullman: Nice discussion of ways intervals contextualize estimates. @Lester_Domes idea to reframe as high compatibility values c…
view full postApril 5, 2022
2
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)RT @JessicaHullman: Nice discussion of ways intervals contextualize estimates. @Lester_Domes idea to reframe as high compatibility values c…
view full postApril 5, 2022
2
-
Jessica Hullman
@JessicaHullman (Twitter)Nice discussion of ways intervals contextualize estimates. @Lester_Domes idea to reframe as high compatibility values conditional on all assumptions https://t.co/Uh7vemZwNx (& other ideas to reform interpretation, w @dailyzad https://t.co/VjyGaGpX9i) should be required reading https://t.co/V9w4zdcttT
view full postApril 5, 2022
7
2
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)@CMastication @rlmcelreath @dylanarmbruste3 @NoahHaber Perhaps, but to be fair, this was for a very specific audience/discussion, he’s written several other papers with more accessible definitions/descriptions of P-values, here are some excerpts from the first link 1. https://t.co/DepyiqagSt 2. https://t.co/b4nrvhCqik https://t.co/lP6IgT9Rg0
view full postApril 2, 2022
4
-
Pavlos Msaouel
@PavlosMsaouel (Twitter)@ESMDcan123 Indeed although unfortunately all statistics is this hard. Bayesian approaches may be more intuitive in many ways but come with their own caveats. One solution may be to prioritize teaching simple cognitive aids such as the s-value transformation
view full postMarch 30, 2022
-
Hind
@guusjink04 (Twitter)RT @dailyzad: @DataSciFact @JohnDCook It’s true, and there are so many other misconceptions regarding what hypotheses you can generate p va…
view full postMarch 3, 2022
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)@DataSciFact @JohnDCook It’s true, and there are so many other misconceptions regarding what hypotheses you can generate p values for and many misconceptions of assumptions. Here’s a definition we give in our paper: https://t.co/uUDid9Iv4T https://t.co/i9FpvzV78w
view full postMarch 3, 2022
29
-
@load-depentent.bsky.social
@load_dependent (Twitter)@oyvlie @drjohnm P-values are if anything much closer to measures of fragility, and if the scale is the hindrance to interpretability, convert them to S-values for an intuitive scale ("more surprising than x heads or tales in a row") https://t.co/xPsN3F0pch
view full postFebruary 20, 2022
-
Jussi Mustonen
@JusiMD (Twitter)@NevalainenOlli @KariTikkinen @markkumyllykan2 @Laakarilehti @Laakariliitto @mediuutiset @yletiede @TampereUniMET @HelsinkiUniMed @JarnoRutanen @AleksiReito Uh, en kyllä tiedä. Oletko lukenut tämän? https://t.co/DftqpTbFD8
view full postFebruary 19, 2022
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)@ENirenberg If you replaced P-values in history (first documented use in the 1700s) with some other statistic that was as widely adopted, I think you would see just the same sort of misuse/misinterpretations. Here's an excerpt from a paper on this topic https://t.co/uUDid9Iv4T https://t.co/Q7bn2AVjps
view full postFebruary 13, 2022
10
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)RT @bmwiernik: @RyanPBowles This thread explains my perspective pretty well I think. And @dailyzad’s article https://t.co/sBxuC0hWtS does a…
view full postFebruary 5, 2022
1
-
Brenton Wiernik
@bmwiernik (Twitter)@RyanPBowles This thread explains my perspective pretty well I think. And @dailyzad’s article https://t.co/sBxuC0hWtS does a nice job of illustrating interpreting a CI as a range of values compatible with the data, model assumptions, and sampling distribution of the estimator https://t.co/1yShSjWUVF
view full postFebruary 5, 2022
1
1
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)@omaclaren @MarkAnnuncio @bmwiernik Along with another example of what it insinuates from our BMC paper : https://t.co/2Zsyh4QtKa https://t.co/FfRfafQxYp
view full postJanuary 17, 2022
2
-
Bob
@fluxcapdoodah (Twitter)RT @IsabellaGhement: @stephensenn @david_colquhoun @learnfromerror @Abraham_RMI @SciReports @Lester_Domes @ken_rothman @goodmanmetrics @vam…
view full postJanuary 14, 2022
1
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)RT @Lester_Domes: @david_colquhoun @vamrhein @dailyzad @learnfromerror David, would you please be clear that our call is to retire deceptiv…
view full postJanuary 11, 2022
9
-
Valentin Amrhein
@vamrhein (Twitter)RT @Lester_Domes: @david_colquhoun @vamrhein @dailyzad @learnfromerror David, would you please be clear that our call is to retire deceptiv…
view full postJanuary 11, 2022
9
-
Sander Greenland
@Lester_Domes (Twitter)@david_colquhoun @vamrhein @dailyzad @learnfromerror David, would you please be clear that our call is to retire deceptive terms like "significance" and "confidence" to describe P-values and intervals based on P-value functions (which show the compatibility of effect sizes with data), NOT to retire P-values https://t.co/Tnq4bZSoAQ
view full postJanuary 11, 2022
17
9
-
Valentin Amrhein
@vamrhein (Twitter)@ExerciseBiology Thanks – not more likely but more compatible. The point estimate is the value that is most compatible with the data and the model, yet it is very unlikely that a point estimate is the true population value. See https://t.co/rV5lJNty7h and https://t.co/DEyXF8tagJ
view full postDecember 18, 2021
1
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)RT @Lester_Domes: @lakens For some review material see Greenland & Rafi https://t.co/8ucXcY7Pjr https://t.co/Tnq4bZANcg https://t.co/MXTzWp…
view full postNovember 23, 2021
3
-
Andrés Fandiño
@andres_fandino (Twitter)RT @Lester_Domes: @lakens For some review material see Greenland & Rafi https://t.co/8ucXcY7Pjr https://t.co/Tnq4bZANcg https://t.co/MXTzWp…
view full postNovember 21, 2021
3
-
Sander Greenland
@Lester_Domes (Twitter)@lakens For some review material see Greenland & Rafi https://t.co/8ucXcY7Pjr https://t.co/Tnq4bZANcg https://t.co/MXTzWp1ecQ Related: Perezgonzalez. P-values as percentiles. https://t.co/yKnsRyi8qM. Vos, Holbert. Frequentist inference without repeated sampling. https://t.co/XwsIfrqYdl.
view full postNovember 21, 2021
6
3
-
Isabella R. Ghement
@IsabellaGhement (Twitter)@stephensenn @david_colquhoun @learnfromerror @Abraham_RMI @SciReports @Lester_Domes @ken_rothman @goodmanmetrics @vamrhein Personally, I found this article helpful towards that goal of thoughtfulness in working with p-values: https://t.co/NJsW4ablRm
view full postOctober 24, 2021
5
1
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)RT @kaz_yos: Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surprise | B…
view full postAugust 16, 2021
3
-
Pavlos Msaouel
@PavlosMsaouel (Twitter)RT @kaz_yos: Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surprise | B…
view full postApril 22, 2021
3
-
WillPer
@wilpertwitt (Twitter)RT @kaz_yos: Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surprise | B…
view full postApril 21, 2021
3
-
Kazuki Yoshida
@kaz_yos (Twitter)Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surprise | BMC Medical Research Methodology | Full Text https://t.co/XQitszfKjX
view full postApril 21, 2021
10
3
-
Karandeep Singh
@kdpsinghlab (Twitter)RT @dailyzad: @kdpsinghlab From our paper (click img to see in full) https://t.co/h21MqyylSO https://t.co/4CCMxJscl3
view full postJanuary 25, 2021
1
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)@kdpsinghlab From our paper (click img to see in full) https://t.co/h21MqyylSO https://t.co/4CCMxJscl3
view full postJanuary 25, 2021
4
1
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)@f2harrell @vandy_biostat @EdgeforScholars Further discussion here: https://t.co/DepyiqagSt https://t.co/E5gzPha7Hh
view full postNovember 19, 2020
4
1
-
Enrique Schisterman
@eschisterman1 (Twitter)RT @dailyzad: @ProfMattFox @eschisterman1 @edwardsjk @nondogmatist @Lester_Domes @LEBiddle great thread! Also see this recent paper which d…
view full postNovember 10, 2020
2
-
Dr. د. عبدالواحد الزهراني
@ralzahrani2020 (Twitter)RT @dailyzad: @ProfMattFox @eschisterman1 @edwardsjk @nondogmatist @Lester_Domes @LEBiddle great thread! Also see this recent paper which d…
view full postNovember 10, 2020
2
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)@ProfMattFox @eschisterman1 @edwardsjk @nondogmatist @Lester_Domes @LEBiddle great thread! Also see this recent paper which discusses P-value functions and S-values and their functions in more depth: https://t.co/DepyiqagSt
view full postNovember 10, 2020
4
2
-
Tilastokunto
@tilastokunto (Twitter)Lukemistoa on päivitetty!
view full postOctober 26, 2020
4
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)RT @dailyzad: @omaclaren @analisereal @kareem_carr @NoahHaber @Andrew___Baker I prefer the P-value as the percentile at which the observed…
view full postOctober 6, 2020
1
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)@omaclaren @analisereal @kareem_carr @NoahHaber @Andrew___Baker I prefer the P-value as the percentile at which the observed test statistic fell in the test distribution per Fraser and Perezgonzalez, which we also mention in our paper: https://t.co/h21MqyylSO https://t.co/lv2IAB59GI
view full postOctober 6, 2020
7
1
-
UCSF Geriatrics
@UCSFGeriatrics (Twitter)RT @wjdeardorff: "Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surpris…
view full postOctober 4, 2020
2
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)RT @IsabellaGhement: @noahaskell @PhDemetri I like this intro to p-value functions: https://t.co/iKPMHhADEb. See also the recent article "…
view full postOctober 3, 2020
1
-
Isabella R. Ghement
@IsabellaGhement (Twitter)@noahaskell @PhDemetri I like this intro to p-value functions: https://t.co/iKPMHhADEb. See also the recent article "Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surprise" by Zad Rafi and Sander Greenland: https://t.co/NJsW4asWIU.
view full postOctober 3, 2020
5
1
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)RT @wjdeardorff: "Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surpris…
view full postOctober 3, 2020
2
-
James Deardorff
@wjdeardorff (Twitter)"Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surprise" (2020) - P-value, S-value, compatibility interval - https://t.co/kCLSE1lHSo https://t.co/JrSWSakILo
view full postOctober 3, 2020
5
2
-
Didier Brassard, PhD
@didier_dtp (Twitter)RT @dailyzad: What even is a P-value tho? (Click image to view in full) https://t.co/h21MqyylSO https://t.co/CoxtyZmLDC
view full postOctober 2, 2020
2
-
Abraham E. Gracia R.
@Abraham_RMI (Twitter)RT @dailyzad: What even is a P-value tho? (Click image to view in full) https://t.co/h21MqyylSO https://t.co/CoxtyZmLDC
view full postOctober 2, 2020
2
-
Luis Mella
@Mella10G (Twitter)@Stfko_memero @mafsant_ @MonicaPerezEro @drfhowley @ImerConceptoz What even is a p-value tho? https://t.co/SZ0bgrx4ch https://t.co/qeg2OJOqPQ
view full postOctober 2, 2020
3
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)@richarddmorey @Lester_Domes If you haven’t already seen, now here: https://t.co/h21MqyylSO
view full postOctober 2, 2020
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)What even is a P-value tho? (Click image to view in full) https://t.co/h21MqyylSO https://t.co/CoxtyZmLDC
view full postOctober 2, 2020
2
2
-
Lucia Ruggiero EALR
@luciaruggiero (Twitter)RT @GlobalPHObserv: Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surpr…
view full postOctober 2, 2020
3
-
Equity and Health
@equityhealthdev (Twitter)RT @GlobalPHObserv: Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surpr…
view full postOctober 2, 2020
3
-
Culture of Health Eq
@CofHEQ (Twitter)RT @GlobalPHObserv: Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surpr…
view full postOctober 2, 2020
3
-
Oliver Maclaren
@omaclaren (Twitter)RT @dailyzad: Glad to see this finally up, it’s open access so anyone can read it and has been in the works for nearly two years, partly du…
view full postOctober 1, 2020
2
-
Sir Panda (Zad Rafi)
@dailyzad (Twitter)Glad to see this finally up, it’s open access so anyone can read it and has been in the works for nearly two years, partly due to the amount of helpful feedback and comments we’ve received. 1/n link: https://t.co/h21MqyylSO enhanced PDF: https://t.co/1CTzHHhBgY https://t.co/Mau8GBd1jw
view full postOctober 1, 2020
12
2
-
Global Health Observ
@GlobalPHObserv (Twitter)Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surprise https://t.co/ITyfqrXnp9
view full postOctober 1, 2020
2
3
-
Ash Paul
@pash22 (Twitter)Semantic and cognitive tools to aid statistical science: replace confidence and significance by compatibility and surprise https://t.co/VkeiUtk6ok via @dailyzad & @Lester_Domes
view full postOctober 1, 2020
Abstract Synopsis
- Researchers often misinterpret statistical outputs, leading to a push for new methods to replace traditional P-values with tools like Bayes factors and compatibility measures.
- Cognitive rather than purely statistical issues are the root of these misinterpretations, prompting the review of methods to better help researchers understand their data with a focus on logical concepts.
- The authors advocate for reinterpreting P-values as measures of compatibility, utilizing tools like the Shannon transform and alternative hypothesis testing to improve clarity and combat common fallacies in statistical analysis.
黒木玄 Gen Kuroki
@genkuroki (Twitter)