@mspro Genau. Das liegt mMn das Problem wie wir aktuell mit dem Begriff der Verschwörungstheorien umgehen (i.e. VTs sind prima facie immer falsch und schädlich). Würde den Zweifel am Selbstmord schon als Verschwörungstheorie werten, nur eben als angemessen begründete.
@strippel Den Thread habe ich heute auch gelesen und fand’s irgendwie traurig.
@mspro Die Unterscheidung ist leider nicht so super trennscharf. In einem Paper von meinen Kolleg*innen und mir schauen wir uns die Social Media Reaktion auf Epsteins Tod an und können ganz gut nachverfolgen, wie nah sich die angemessene Skepsis und das unbegründete Q-Schwurbeln sind.
@chainsawriot Auch ein merkwürdiges Wort: „doch“
yourdailygerman.com/meaning-of-doc…
Playing around with #v2 of the @TwitterDev API. Looks promising!
To the #rstats/#nlp people in my timeline:
Are you like me and always get anxious seeing the deprecation warnings when using widyr::pairwise_count(), hoping it won't break your scripts when dplyr get's updated?
If that's the case, take a look at this: https://t.co/lM8YLHvcdW 
It's a drop-in replacement of the widyr function and is even a little faster / needs a little less memory on large datasets. Feel free to copy the code to ease your deprecation anxiety.
@Dede198904 @CogitatioMaC @OSFramework Yeah, more like fauxpen data. But i guess that deleted content on any social media platform is a tough issue for #openScience. From a technical and an ethical perspective.
4/4
So, if you are in #entertainmentResearch and would like to get started with #computationalMethods, give our paper and the whole SI a read and check out our scripts on the OSF. The scripts are released under the MIT license (opensource.org/licenses/MIT).
osf.io/vzymj/
2/4
As computational methods (such as automated content analysis) are relatively new to #entertainmentResearch, we wanted to showcase how these methods work and how researchers can apply them in their own work. So we added all our final scripts to OSF repo to facilitate this.
3/4
Sadly, as per the Twitter TOS, we cannot share the complete dataset freely, so we have shared the list of Twitter status IDs that can be rehydrated via the API. It's important to note that, because we analyzed historical data, a large share of Tweets will be unavailable. 
1/4
As our current paper in the @CogitatioMaC special issue "Computational Approaches to Media Entertainment Research" is not only #openAccess, but also #openCode and #openData (kinda), I wanted to talk about the @OSFramework repo here as well.
twitter.com/Kudusch/status…
You can find the paper and all the other great work in the special issue here: cogitatiopress.com/mediaandcommun…
@d_stoltenberg @WHO Thank you! I usually look at the top documents by topic and try to chose a quote that fits well (both in terms of content and length). It's more of a gut-feeling/qualitative approach :)
We analyzed the debate on #GamingDisorder and #GamingAddiction on Twitter, following the @WHO's decision to include "Gaming disorder" in the ICD-11. We used a mix of co-occurrence graphs, sentiment analysis, and topic modeling. 
This paper is part of a great special issue on #ComputationalMethods in #EntertainmentResearch, edited by @MattEagle09, @WimTulf, & @romohseni (who were the most constructive editors!)
twitter.com/WimTulf/status…
➡️📃New Paper!📃⬅️
"A Computational Approach to Analyzing the Twitter Debate on Gaming Disorder" by Schatto-Eckrodt, Janzik, Reer,
@ZwiZwaSvens & @thorstenquandt.
The paper is #OpenAccess and the analysis scripts are #OpenSource, so take a look! 
@mspro Genau! Das passt super zu der Wood, Douglas, & Sutton, 2012 Studie. Es gibt mit Sicherheit einige tatsächlich Überzeugte in der Gruppe der VTler, aber bei denen, die nur performativ zustimmen, bringen korrigierende Informationen auch nichts.
@mspro Diese Argumentation hat mir in meiner Forschung zu dem Thema weitergeholfen: Nur weil jemand *sagt*, dass sie an eine Verschwörungstheorie glauben, heißt das nicht, dass sie *wirklich denken*, dass sie wahr und real ist.
https://t.co/hcmzdqQHVt 
@monkeydom sort -R does something similar. It shuffles but groups duplicates.
@Dede198904 Ist das ein Thema der beliebten wissenschaftlichen Disziplin „EDV“?
@mariavosk Ich glaube fast, da ist bei der Übersetzung einer Pressemitteilung was schief gelaufen 🙃
Q-Anon? Wirkmächtiges Verschwörungsnarrativ, das mittlerweile auch in Deutschland fußgefasst hat. Bei @Frontal21 gibt es dazu eine gut gemacht Doku (~ 17 Minuten). Sehr sehenswert, unter anderm wegen dem Auftritt von @_pietz_ und Musik von @Fatonimusik!
twitter.com/felixhuesmann/…
Ich verstehe weder was „Infodemiologie“ sein soll, noch warum das eine neue Disziplin darstellen soll. In der Meldung steht ja, dass Wissenschaftler*innen aus „Mathematik, EDV(!), Soziologie, Psychologie, Gesundheit und Kommunikation“ schon dazu forschen.
zeit.de/wissen/gesundh…
This is a pet-peeve of mine. While this specific example probably is meant as a stylistic device, just slapping a handful of citations on a sentence won't help me understand your reasoning.
This is also a good example how citation style will affect academic writing style. 
@Lenafrescamente Jo, sieht spanned aus!
@mariabloec In the groups I worked this was always the first-authors job, regardless of hierarchical status. I don’t know how I would react if I was only a co-author and had to to the copy-paste work :D
Web Design Museum exhibits over 1,600 carefully selected and sorted web sites that show web design trends between the years 1991 and 2006.
https://t.co/KKe2QmEzwk 
@bowmanspartan @WY_Tang @Twitter You’re welcome! Exactly, this was part of a @ica_cm panel.
Im @_forschergeist_ Podcast spricht @timpritlove mit @JanSchmidt über Blogs, den Medienwandel und die Deutungshoheit traditioneller Medien.
Schönes Stück #Wissenschaftskommunikation darüber, was mein Fach die #Kommunikationswissenschaft so u.a. macht.
forschergeist.de/podcast/fg080-…
@cbpuschmann Although your Sentiment section is already pretty extensive, the sentimentR might be worth a look! It takes valence shifters into account, has cool utility functions (emojis replacement!), and an interesting workflow.
github.com/trinker/sentim…
@sjacks26 Thank you! The virtual ICA presentation is still online (player.vimeo.com/video/417213256) and the software itself is online as well (github.com/Kudusch/twitte…). If you have any questions, feel free to ask!
… oder ihr schaut euch die Studie selbst an, #openaccess auf #arXiv:
Teil 1: arxiv.org/abs/2004.02566
Teil 2: arxiv.org/abs/2005.13290
Unser zweites Working Paper über die mediale Berichterstattung während der #Coronakrise ist letzte Woche erschienen! Was @thorstenquandt, @ZwiZwaSvens, @Lenafrescamente, und ich herausgefunden haben, könnt ihr hier nachlesen: https://t.co/dkXovCG1Vt 
4/5: Last but no least: "Hidden Biases: The Effects of Deleted Content on Twitter on Sampling Quality". Here we investigated how Twitter samples deteriorate over time and how this affects sampling quality. The likelihood of deletion grows with age and is higher for certain topics 
5/5 All this research was conducted by my co-authors (in various configurations): @ZwiZwaSvens, @Lenafrescamente, @thorstenquandt, @Flo_Wintt, L. Clever, and me.
If you are attending the virtual #ICA20, look for our presentations and if you have any questions, feel free to @ me!
3/5: Related to that, in our presentation "The EU Election 2019 on Social Media" we compared alternative and established news media. We find that established media had a much larger reach than alternative media, yet alt. media posts often reach higher average number of likes. 
Thread, 1/5: Now that #ICA20 has been extended by a day and most presentations are online and at the right place, I'll give you a (very condensed) peak at what my co-authors and I submitted this year (session title and division in the screenshots!):
2/5: In our presentation "Denunciation, Corrective or Catchphrase? The Use of the Term Fake News in Social Media Debates" we find that the term "fake news" is used mainly to attack established media outlets in many different media systems. 
@chainsawriot Sadly is it! I have contacted the ICA staff on Wednesday but haven’t heard back from them. I can send you the slides/the video later today if you’re interested!
@Lenafrescamente @ryanjgallag @JoeWasserman @ZacharyST @p_barbera @ecozzo As @Lenafrescamente said, I am working on a paper looking at the effects of deleted content on sampling quality and have lots of data on the topic (both decahose and gardenhose).
Sneak peak at the data: 
@jasonbaumgartne Relevant Geek and Poke comic. 
@hase_va @chainsawriot It is! And it’s really easy to use as well.
This is an important step more social media companies should take: blog.twitter.com/developer/en_u…
Addendum: The peak in late February corresponds to the terrorist attack in #Hanau, where the shooter's manifesto mentioned the "great replacement" conspiracy theory.
While it'll take time to gather enough empirical data to conclusively answer the question if #COVID19 gave rise to more #ConspiracyTheories, we definitely seem to talk more about the topic. Both the number of tweets and users mentioning "conspiracy theories" are on the rise. 
@Dede198904 @affinitybyserif Leider nicht, nee. Keine Ahnung ob sowas in Planung ist.
Falls jemand von euch überlegt hat von #Adobe zu den @affinitybyserif Produkten zu wechseln, jetzt ist der Moment! 50% Rabat auf alle Apps (Windows, Mac, iOS).
Richtig gute Software ohne Abo-Model. Hat für mich Illustrator, Photoshop und InDesign restlos ersetzt.
@hase_va Right?! Connecting domain-specific knowledge and automated methods is such a cool approach.
This looks really interesting: Keyword Assisted Topic Models (keyATM), which combines LDA models with a small number of keywords selected by researchers in order to improve the interpretability and topic classification of the LDA.
arxiv.org/abs/2004.05964