@katiedimartin @NicholasTwiner And yet, is has been tried several times again... I also was gonna mention that pro-drop paper, but there's even more: https://t.co/MI5OVhuAfQ The 'linguistic diversity and traffic accidents' paper should be compulsory readin
@TomPepinsky I'm finding it surprising that you haven't cited Winters and Roberts extensive work on how to overcome spurious correlations between linguistic x extralinguistic vars (given that you cited the Chen study) E.g. https://t.co/n7QVL3qjVX And more
@Zupancia Right -- people have also found a statistical correlation between tone languages and the presence of acacia trees, or linguistic diversity and road traffic fatalities, but that doesn't mean the correlations are nonspurious https://t.co/fpa12nnmW7
RT @matspike: I think the first issue, i.e. complaints regarding technical merit, are business as usual and good, *although* it's worth not…
I think the first issue, i.e. complaints regarding technical merit, are business as usual and good, *although* it's worth noting that the authors explicitly address many of the issues & are *very very* aware of problematic causal claims, like *very*⬇️⬇
RT @PsychoSchmitt: "The probability of a spurious correlation between any two cultural traits is higher than is sometimes appreciated by re…
"The probability of a spurious correlation between any two cultural traits is higher than is sometimes appreciated by researchers, the media and the general public." https://t.co/WxSAzb3Wj6
@Ted_Underwood @benmschmidt @scott_bot @LeoImpett Exactly! Post hoc spurious correlations are the main danger of any large-data study – DH included. Here's a fun & useful paper by Sean Roberts & @replicatedtypo on this matter https://t.co/SRWD93Voe
@acerbialberto We also found that chocolate consumption correlates with the number of serial and rampage killers 😉 https://t.co/MJyyKk6GR6
RT @jawadywn: context: https://t.co/H4Zwszb07n
context: https://t.co/H4Zwszb07n
@LinguaDigitalis @MatteoFuoli @aucksAdventures @replicatedtypo Ohhh, if you like this stuff, check out this paper: https://t.co/LwJdKwb1bn Some of the gem "findings" include: Languages spoken by people who take siestas have less complex verb inflection, an
Sean Roberts @aucksAdventures talks about his work with James Winters @replicatedtypo on spurious correlations - read the paper if you haven't done so yet https://t.co/iUrA0tzHBo #SALC7
RT @SimonKirby: @shravanvasishth I think @aucksAdventures and @replicatedtypo ‘s brilliant paper (from which this lovely figure is taken) s…
@shravanvasishth I think @aucksAdventures and @replicatedtypo ‘s brilliant paper (from which this lovely figure is taken) should be required reading before anyone looks at a cross-cultural dataset. :-) https://t.co/BzZvq65kDx
RT @bioculturalCSU: At #CESconf2018 repeatedly struck by the dangers of spatial autocorrelation, phylogenetic autocorrelation and spurious…
RT @bioculturalCSU: At #CESconf2018 repeatedly struck by the dangers of spatial autocorrelation, phylogenetic autocorrelation and spurious…
RT @bioculturalCSU: At #CESconf2018 repeatedly struck by the dangers of spatial autocorrelation, phylogenetic autocorrelation and spurious…
RT @bioculturalCSU: At #CESconf2018 repeatedly struck by the dangers of spatial autocorrelation, phylogenetic autocorrelation and spurious…
At #CESconf2018 repeatedly struck by the dangers of spatial autocorrelation, phylogenetic autocorrelation and spurious correlations. Reminded many times about important issues previously highlighted nicely by @aucksAdventures @replicatedtypo: https://t.co/
@linguatical Mmm, I don't have any dubious papers to tell you in public, but @aucksAdventures and @replicatedtypo have a great paper about the use of correlational studies in linguistics more generally: https://t.co/6sWoB5wgre
@mikejonesmelb #DHA2016 And this one: https://t.co/vP1EFvSRTH
Muito bom: "Linguistic Diversity and Traffic Accidents: Lessons from Statistical Studies of Cultural Traits" https://t.co/4PCz7wB2L1
Linguistic Diversity and Traffic Accidents: Lessons from Statistical Studies of Cultural Traits http://t.co/MdQBjayE1y
Lots of #spuriouscorrelations in the news (http://t.co/bHohnbz6cs), so here's some science about why they appear http://t.co/p3CSlnzMY5
@MikeYearworth I think you will like this paper; it look at spurious correlations in data: http://t.co/LHfGXEPZ7f
@MikeYearworth I think you will like this paper; it look at spurious correlations in data: http://t.co/LHfGXEPZ7f
@nikhilnarayanan These correlation studies are pure funny. Also, http://t.co/7GCxsog0La
今日のTSに関連して、言語の特徴と文化の関係について、いくらでも怪しげな相関を作れるので注意しましょうねという論文: http://t.co/nt3RPQjdQ7
今日のTSに関連して、言語の特徴と文化の関係について、いくらでも怪しげな相関を作れるので注意しましょうねという論文: http://t.co/nt3RPQjdQ7
今日のTSに関連して、言語の特徴と文化の関係について、いくらでも怪しげな相関を作れるので注意しましょうねという論文: http://t.co/nt3RPQjdQ7
今日のTSに関連して、言語の特徴と文化の関係について、いくらでも怪しげな相関を作れるので注意しましょうねという論文: http://t.co/nt3RPQjdQ7
@CaMLA_Global no no no no no no http://t.co/tFY9e4E9JQ http://t.co/WMButkNUT5
@taaljournalist Tja. Uit een hoop data zijn altijd wel een paar correlaties te schudden. Zie hier: http://t.co/WFIVTSvBoL
Boobs http://t.co/jCdQayexeD
Waanzinartikel: het verband tussen alles. http://t.co/Ha7jg8xCgq
Pithy new study on false correlations: http://t.co/MGpEF5naan
Great, snarky @PLosOne study on correlation vs. causation: http://t.co/SRvFNHXdgP And @Slate's take: http://t.co/0ogstcvX0b cc: The Media
Great, snarky @PLosOne study on correlation vs. causation: http://t.co/SRvFNHXdgP And @Slate's take: http://t.co/0ogstcvX0b cc: The Media
Big data double win: "revised statistical standards" would accept sex-traffic-trees #oops http://t.co/O13Dscx1tr http://t.co/mH6GZrLKyj
Big data double win: "revised statistical standards" would accept sex-traffic-trees #oops http://t.co/O13Dscx1tr http://t.co/mH6GZrLKyj
Big data double win: "revised statistical standards" would accept sex-traffic-trees #oops http://t.co/O13Dscx1tr http://t.co/mH6GZrLKyj
Big data double win: "revised statistical standards" would accept sex-traffic-trees #oops http://t.co/O13Dscx1tr http://t.co/mH6GZrLKyj
Great, snarky @PLosOne study on correlation vs. causation: http://t.co/SRvFNHXdgP And @Slate's take: http://t.co/0ogstcvX0b cc: The Media
Correlation vs. causation http://t.co/awCnoVOmgh in @PLosOne http://t.co/R7gXdpp1av in @Slate via @DavidJuurlink #journalism
"spurious correlations are much more likely than some studies admit" in #PLOSONE: http://t.co/lMHPL285fN
RT @picardonhealth: Correlation vs. causation http://t.co/3UFlKLqnqR in @PLOSONE http://t.co/ei5uE4BHGm in @Slate via @DavidJuurlink
Great, snarky @PLosOne study on correlation vs. causation: http://t.co/SRvFNHXdgP And @Slate's take: http://t.co/0ogstcvX0b cc: The Media
Correlation vs. causation http://t.co/awCnoVOmgh in @PLosOne http://t.co/R7gXdpp1av in @Slate via @DavidJuurlink #journalism
Correlation vs. causation http://t.co/awCnoVOmgh in @PLosOne http://t.co/R7gXdpp1av in @Slate via @DavidJuurlink #journalism
Correlation vs. causation http://t.co/awCnoVOmgh in @PLosOne http://t.co/R7gXdpp1av in @Slate via @DavidJuurlink #journalism
.@joshuakeating piece in last tweet highlights great critique by @replicatedtypo & @aucksAdventures on correlations: http://t.co/q3btCwerxi
Great, snarky @PLosOne study on correlation vs. causation: http://t.co/SRvFNHXdgP And @Slate's take: http://t.co/0ogstcvX0b cc: The Media
Great, snarky @PLosOne study on correlation vs. causation: http://t.co/SRvFNHXdgP And @Slate's take: http://t.co/0ogstcvX0b cc: The Media
Great, snarky @PLosOne study on correlation vs. causation: http://t.co/SRvFNHXdgP And @Slate's take: http://t.co/0ogstcvX0b cc: The Media
Enamoured with 'big data'? Here's a humorous and scholarly reminder of its perils (PLoS One) http://t.co/xucofKZ2uX via @DavidJuurlink
Great, snarky @PLosOne study on correlation vs. causation: http://t.co/SRvFNHXdgP And @Slate's take: http://t.co/0ogstcvX0b cc: The Media
Great piece of correct academic snark partially by @UniofEdinburgh person @replicatedtypo http://t.co/B7yjOYUBac
This paper attempts to link the most random things to show how people are linking the most random things. http://t.co/5z2L9ZbpFp
@Inoueian http://t.co/toABEdE3pw 論文。PLOS ONE(←全部大文字になったのね)だからタダで読める。Figure 3とか良い。
The problems with correlations http://t.co/QOgx8NDWLZ
Linguistic Diversity and Traffic Accidents: Lessons from Statistical Studies of Cultural Traits http://t.co/VWF6YvK8a9
100% of Euro-Canadians who eat tomatoes die: spurious cross-cultural links in nomothetic studies #NotByTheNumbers http://t.co/mdIaUmedoC
Study: pretty much everything is correlated with pretty much everything else http://t.co/2OwcvXm55N via @Slate
Rinse, repeat. Correlation ≠ Causation. Everyone u know must read this. Linguistic Diversity and Traffic Accidents. http://t.co/lDFnzEqB84
The problems with correlations http://t.co/QOgx8NDWLZ
The problems with correlations http://t.co/QOgx8NDWLZ
Gave studie van collega Seán Roberts: "Linguistic diversity and traffic accidents", ofwel correlatie ≠ causatie http://t.co/WFIVTSvBoL #taal
In @PLOSONE, @aucksAdventures and @replicatedtypo do for correlational studies what the dead salmon did for fMRI: http://t.co/msFPYSpjZ3
Suklaan syöminen ei sitten ennustanutkaan nobeleita #tiede #suklaa http://t.co/VvSBpTkqHz
This is a great critique of many large-scale correlational studies http://t.co/AmsrGcvCKG @aucksAdventures @replicatedtypo
"Linguistic Diversity and Traffic Accidents: Lessons from Statistical Studies of Cultural Traits" http://t.co/Hfk80jMoDa
This is a great critique of many large-scale correlational studies http://t.co/AmsrGcvCKG @aucksAdventures @replicatedtypo
This is a great critique of many large-scale correlational studies http://t.co/AmsrGcvCKG @aucksAdventures @replicatedtypo
This is a great critique of many large-scale correlational studies http://t.co/AmsrGcvCKG @aucksAdventures @replicatedtypo
This is a great critique of many large-scale correlational studies http://t.co/AmsrGcvCKG @aucksAdventures @replicatedtypo
This is a great critique of many large-scale correlational studies http://t.co/AmsrGcvCKG @aucksAdventures @replicatedtypo
This is a great critique of many large-scale correlational studies http://t.co/AmsrGcvCKG @aucksAdventures @replicatedtypo
This is a great critique of many large-scale correlational studies http://t.co/AmsrGcvCKG @aucksAdventures @replicatedtypo
Linguistic diversity is correlated with traffic accidents! Why you should be wary of big data studies: http://t.co/SxK0CvMPKz
@nntaleb Spurious statistics everywhere! Acacia trees predict traffic accidents, which predict linguistic diversity. http://t.co/2iwzAC5jIP
“This is the tragedy of big data: The more variables, the more correlations that can show significance." http://t.co/alpU7yhN3N
RT @aucksadventures: Linguistic Diversity & Traffic Accidents.Oh, & eating choc ->serial killers! http://t.co/BpP5GviXrO
@nntaleb Spurious statistics everywhere! Acacia trees predict traffic accidents, which predict linguistic diversity. http://t.co/2iwzAC5jIP
This is a great critique of many large-scale correlational studies http://t.co/AmsrGcvCKG @aucksAdventures @replicatedtypo
Be careful out there, folks! >> Linguistic Diversity and Traffic Accidents: http://t.co/17NDa5643u #linguistics #bigdata #correlation
@nntaleb Spurious statistics everywhere! Acacia trees predict traffic accidents, which predict linguistic diversity. http://t.co/2iwzAC5jIP
Linguistic Diversity and Traffic Accidents: Lessons from Statistical Studies of Cultural Traits http://t.co/PyfPcVK7yw #fb
En este estudio x ej exploran asociación e/ accidentes de tráfico y la diversidad lingüística de una población http://t.co/DIuJP5MxOz
@deevybee Our rejected stats on Chocolate and Nobel Laureates now published. Eating choc leads to serial killers! http://t.co/Ad11YrwIC8
“This is the tragedy of big data: The more variables, the more correlations that can show significance." http://t.co/alpU7yhN3N
Linguistic Diversity and Traffic Accidents: Lessons from Statistical Studies of Cultural Traits http://t.co/vVGEdTEH3u