Go forth Jitka’s immune system!
2 Likes
Dear All, I added my minutes of our April 16 meeting (with Richard’s comments) to our evolving document. I need to do prepare one more part, about the approach I am developing to populism as a cultural phenomenon. Will try tomorrow - got absorbed by other duties. Just in case: the document is here:https://docs.google.com/document/d/1AqmP5sM1wPzHVi7vF6HT1LnL3gq9XGcjKkh0Px1Vk_c/edit
1 Like
Dear @rebelethno and @alberto
Question for the hive mind:
While Djan and I wait for UCL ethical approval for the ethnographic interviews, he will be putting together a corpus of online posts by the far-right populist Attila Hildmann. We want to make sure that we cannot be accused of cherry-picking particularly outrageous posts, so can anyone recommend any literature on sampling online sources? Is there anything we need to bear in mind that’s different from sampling newspapers, for instance?
1 Like
For some context for all the deepl/Google translate aficionados
I think in that case you are studying discursive practices and narrative constructs, so representatives is not the goal itself - I would say, you are studying f.e. his practices of othering, his idea of “we” vs “them”, and in that regard, you can base your analysis on a particular timeframe, like a month or two months of his posts, to be the base of the dataset. And then you do a textual/visual analysis of his activity and focus on particular motives, choosing them f.e. on the basis of their popularity, or quantity, like “it’s a majority of his posts”. It isn’t necessary to be representative of his overall activity if you just focus in your analysis on the ways he narrates certain topics if you have this particular time frame as a point of reference (but correct me if I am wrong). I would refer here to basic writings on CDA, aka critical discourse analysis method.
3 Likes
Let me know if you need literature on that 
@Maniamana thanks! Was thinking the same thing re limiting the research focus to a specific time frame. I specifically want to look at voice messages Hildmann posts in his channel. This enables us to transcribe and code them as well. (I just have to convert .ogg to mp3 for the transcription software we are using).
He posts a lot of voice messages a day, and often very lengthy ones. I think it could make sense to randomize which voice messages we transcribe (for example first and last voice message posted on three selected days of a week)
Re focus and compatibility with overarching POPREBEL angle, I was thinking of looking at Hildmann’s antisemitic + homophobic framing of sociopolitical elites through the lens of thick form of populism a la Godfather Cas Mudde
1 Like
Cool, great idea!! Not sure @ the randomization, but if there is really a lot of messages per day, it’s hard to choose. If it is possible though to listen to all of them and choose those that contains the topics of interest, I think it is totally methodologically correct. I don’t suppose he talks only antisemitic and homophobic content xD
1 Like
I would say the majority of his „discourse“ revolves around antisemitic framing. His latest conspiracy myth is that Yehudah Teichtal (community Rabi of the Jewish Community in Berlin) is the real chancellor of Germany.
All political elites he talks about are either Jews or controlled by „Jewish interests“ according to him.
He has a very simplistic Us (the righteous and morally intact) vs. Them (Jews, and their allies) worldview. He projects it into almost everything he talks about.
Agree with Mania here! Time frame is a great way to bound the analysis for your purposes.
2 Likes
@alberto could possibly also think of ways to do more quantitative measures on the posts to pair with your analysis.
2 Likes
@rebelethno, I will be about 30 minutes late to our 2pm UK meeting today. Please feel free to start without me.
For me, we can start at 2.30 (UK time).
@Djan, I would do two things.
The first: add to the posts by Hildmann a Discourse tag, for example #ethno-hildmann. With this, it is easy, both in Tulip and in the future multi-tenancy Graphryder, to induce the SSNA of that particular group of posts.
The second: add the #ethno-hildmann posts to the ethno-poprebel corpus. We have a way to reduce the bias implied by an over-representation of Hildmann’s thinking in the corpus, based on a “one person one vote” logic: an edge in the codes co-occurrence network is only counted once per informant. When we reduce the network, we filter it, so that only edges representing associations made by, say, 4 or more different people remain. This way, associations that are made by Hildmann alone are filtered out: those that remain are those that he shares with others. The technique is explained here.
3 Likes
@Djan @Richard For a very similar time-scope and task, we used constructed weeks (European elections campaigns), constituted randomly, with a smaller control sample of non-electoral period in DEMOS (p. 14-18). However, I agree with Mania , I personally do not think you have to be as rigid when your intention is to study already identified specific narratives and you are approching it through CDA,
The other way out might be stating a relevance-based sampling was used, because you knew you will find it there and quantification was not the guiding principle:)
3 Likes
Dear all
I’ve got six hours of meetings at UCL tomorrow 
, so I won’t be able to make the first day of the hackathon but I’ll pop in on Thursday if that’s OK. Good luck!
R
1 Like
@amelia and All. Are we meeting tomorrow (Friday, 30 April)? I was not able to do much this week or attend the event. I got my second Moderna shot on Tuesday (in itself a half a day event) that knocked me down. Today I am much better, but had to catch up on several neglected tasks. So, I have not worked on our stuff yet.
Yes, we will be online! Check in with us if you can 
Dear all, I just entered my coding view and basically, all my codes have disappeared from my account (-creator: Jirka_Kocian)…did the dissociation happen because of the current tinkering for the purpose of the hackathon? does anyone else have this problem as well?
Hm this is weird. I wonder if it’s because they are now all under categories which I created. The original author of each code is still present, though…