22.02.19 On weapons and dialects: The initial hypothesis and the final story Marie-José Kolly | NZZ Visuals @mjkolly R at the NZZ: Data-driven stories R at the NZZ: Rapid response 1
22.02.19 R at the NZZ: Elections The outcome was unexpected Case 1: on weapon exports • No specific anticipation • Data made a clear point Case 2: on dialects • Expected specific effects • No confirmation The outcome was unexpected The outcome was unexpected Case 1: on weapon exports Case 1: on weapon exports • No specific anticipation • No specific anticipation • Data made a clear point • Data made a clear point Case 2: on dialects Case 2: on dialects • Expected specific effects • Expected specific effects • No confirmation • No confirmation 2
22.02.19 On weapon exports: Motivation Data on weapon exports State Secretariat for Economic Affairs (Seco) SRF Data Debate over export regulation. But how much war material is already sold to warring countries? Hypothesis: No specific anticipation. Data on conflicts Data on conflicts 3
22.02.19 Scripting: Rename countries via countrycode Scripting: «Reshape» data Scripting: Merge dataframes 4
22.02.19 More scripting: visualization Visualisation R to design 5
22.02.19 We write about methods. And publish code on github.com/nzzdev/st-methods The outcome was unexpected On dialects Case 1: on weapon exports Stereotype: «Mischdialekte» • No specific anticipation Hypothesis: • Data made a clear point More mobility => more «mix» More linguistic variability in larger places? Case 2: on dialects • Expected specific effects • No confirmation 6
22.02.19 On dialects Data on dialect use: Crowdsourcing Stereotype: «Mischdialekte» Hypothesis: More mobility => more «mix» More linguistic variability in larger places? Dialäkt Äpp (by. Leeman, A., Kolly, M.-J., Wanitsch, D., Hasler, F., Eisenblatt, C., Werlen, I. Data on dialect use Data on demographics etc. Many, many lines of code that clean data, reduce phonetic categories etc. 7
22.02.19 Statistics: measure linguistic variability via qualvar Deviation from the mode = 0 Deviation from the mode = 1 Results Find an alternative spin 8
22.02.19 Pick extrema pars pro toto Visualize what differentiates them Fribourg Baar Ein paar Grosseltern: Bütschgi Ein paar Enkelkinder: Bütschgi Grosseltern: Gräutschi Grosseltern: Bütschgi Enkelkinder: Bütschgi Ein paar Enkelkinder: Gräutschi Ein paar Grosseltern: Gröibtschi Work on design Do reporter work 9
22.02.19 Explain linguistic change via protagonists This story is still data-driven! 10
22.02.19 Thank you for your attention! Questions? Contributors to the article on weapons: Authors: Marie-José Kolly, Anna Wiederkehr, Michael Surber, Andreas Rüesch, Daniel Steinvorth, Andres Wysling, Patrick Zoll Participation: David Bauer, tools: Beni Buess, Manuel Roth, Philip Küng. Contributors to the article on dialects: Authors: Marie-José Kolly, Alexandra Kohler, Stefanie Hasler (Video), Anna Wiederkehr (Grafik), Annick Ramp (Fotografie). Participation: David Bauer (Konzept), Markus Stein (Video), Beni Buess und Philip Küng (interaktive Elemente). Help and feedback by many others. 11
Recommend
More recommend