Out of 's recent 5,000 followers, over half were created in the past week
1278 Thu Dec 26
647 Fri Dec 27
212 Wed Dec 25
114 Tue Dec 24
89 Mon Dec 23
60 Sun Dec 22
33 Sat Dec 21
32 Wed Dec 18
31 Thu Dec 19
30 Fri Dec 20
Conversation
It's going to be REAL interesting once I get all 68 million of his followers to plot when the accounts were created. Something fishy seems to be going on here.
5
2
31
Not necessarily Jason. Big data should be used carefully, with modeling and causal reasoning. The way Twitter is designed might be responsible for that pattern, as well as some fishy operation, difficult to disentangle all effects. Deeper analysis definitively needed.
1
What do you mean by "casual reasoning?" I am not drawing any conclusions but merely stating a fact that warrants more investigation. As far as I am aware, no one has published analysis on this (or at least I haven't found any).
There are publications on this, because indeed is a good idea! And it's causal, not casual: causal reasoning is a scientific method and helps to avoid to give relevance to patterns that might indeed be spurious or due to something one is not taking into account.
Since the topic is relevant for society as a whole, the role of a scientist is to avoid reporting facts which are not scrutinized through scientific method. Reporting all patterns might be dangerous, because people often tend to amplify some content.. (echo chambers).
1
What you suggest has been firstly reported by a while ago and since then researchers started to think about it. Nevertheless, nobody disentangled all the intervening factors, therefore research is ongoing. Data as the ones you promised would help for sure.
1