Data as Menetekel 7: More than a quarter of the tweets on the migration pact Social Bots?

On 10.12.2018 the WELT published an article by Jan Lindenau: Robots mobilize against migration pact (behind the pay barrier). According to the article, Botswatch evaluated an analysis of 800,000 tweets between 24 November and 2 December. As a result, it announces that over a quarter of these tweets were social bots that created a mood against the migration pact in terms of content. This news was spread throughout the German press.
However, the analysis itself and the methodology proclaimed as a trade secret were not disclosed. This and the incomprehensible results met with considerable criticism from experts. Social media analyst Luca Hammer criticized the statements on Twitter. In a sample, he found that the proportion of bots – including automated information bots from the press – is around 6 percent. Data journalist Michael Kreil has written an open letter to Botswatch:
“We urgently need scientifically based, methodologically correct analyses of these processes! What we don’t need are actors who kidnap the discourse with unproven allegations, spread panic and uncertainty, and inadequately advise the federal government. For me, it is currently not possible to tell which position you would like to take with botswatch. Therefore I ask you, Mrs Wilke, to release the methods and data of your study “Social Bots and Migration Pact” for an independent, scientific review.”

The course of the discussion is described in the articles by Robert Tusch: Kritik an Botswatch: Warum die Debatte um die Social Bot-Studie zum Migrationspakt für Medien ist wichtig, in: Meedia vom 12.12.2018 sowie von Markus Reuter: Ein Bot allein macht keine Revolte. And also no migration debate, reproduced in: netzpolitik.org of 10.12.2018. It also deals with the methodological problems of identifying bots and other forms of accounts, which are more important for political influence on the net. “Alone in view of the influence of humanly operated accounts, it is simply dubious and subcomplex to try to explain social movements or hard-fought social discourses with bots.”
Jonas Hermann: Did bots have a one-sided influence on the debate on the migration pact? Neue Zürcher Zeitung of 13.12.2018 summarises the criticism and discussion once again. He points to Botswatch’s close personal ties to the CDU.
“The bot analysis was published on the day the migration pact was passed. This may or may not be a coincidence. What is certain, however, is that it can be used to discredit critics of the Pact and to present the debate about it as inflated and externally controlled.”
Thus, it is possible that it is not social bots, but freely invented data about social bots that should influence political discourse.