Please disable Adblockers and enable JavaScript for domain CEWebS.cs.univie.ac.at! We have NO ADS, but they may interfere with some of our course material.

Q4 Answers

• In the book it is not really mentioned, that we are better in noticing vertical length than noticing horizontal length, is that true? If yes, why isn't it mentioned/focused?  

• Is there research and development of finding new channels & marks? Or is it proven that all were found?  

• discriminability: what to do, if there are not enough bins?  

• popout: why is it not possible on multiple channels?  

• how has the effectiveness of different channels been explored? what did the experiments look like? 

• Ohne dem gelehrten Hintergrund in der Literatur ist es in meinen Augen klar bzw. natürlich, welche Channels bei welchen Attributes zu verwenden sind. Somit interessant, dass unsere menschliche Wahrnehmung automatisch das passende "Encoding" für gewisse Daten verwendet!  

• Grafik 5.8 von Cleveland and McGill bzw. Heer and Bostock bzgl. der Error Rates ist mir leider nicht ganz klar. Ich habe länger versucht sie zu verstehen aber leider ohne Erfolg, vl. können wir in der Vorlesung kurz darauf eingehen?  

• Unsere Wahrnehmung lässt sich sehr einfach durch Grautöne bzw. generell durch falsche Farbwahl stören. In dem Video kommt sehr schön heraus, dass man durch falsche Farbgebung extreme Verwirrung und auch falsche Ergebnisse erzielen kann. Umso erstaunlicher finde ich es, dass viele Visualisierungen die man in den Medien sieht oft sämtliche dieser Punkte verletzen. Somit kein Wunder, dass viele Visualisierungen komplett falsche Daten vermitteln.  

• Is there a difference how strong we detect identity and magnitude?  

• Differs the mental model of users also on their sex, education, job or other factors?  

• Are there forms that popout more than others or is just the color the most important popout factor?  

• How is a barchart defined only through marks and channels?  

• in 5.4.1. How is the "ordered data" actually to be seen as ordered, if not obviously like ascending etc..  

• I don't understand the law of Stephens. How is the perceived sensation quantifiable - respectively the physical intensity?  

• How can one combine expressiveness and effectiveness to create a ranking of channels according to the type of data?  

• What's the best way to choose which attributes one should encode with position?  

• What kind of other channels do support popout and why isn't possible to use popout with multiple channels?  

• Bezogen auf die Marks und Channels: gibt es eine gewisse maximale Grenze an Marks die man verwenden kann, damit die komplette Grafik noch verständlich bleibt?  

• Wenn man Daten am Besten immer sortiert, wie verhält sich das dann mit Diagrammen, die manipuliert werden können (d3, TAbleau) ist es dort auch von Vorteil automatisch zu sortieren, oder irritiert dies nur?  

• Wenn die menschliche Wahrnehmung eher auf relative Verhältnisse reagiert, begünstigt dies dann den Lie-Factor? 

• Wenn die Daten für den Benutzer rießig wäre, was würde dann bei Visual Encoding passieren? Welche Aufgaben hätte man eine effektive Visual Encoding zu realisieren?  

• Wofür steht dieser "n" in "Law of Stevens" ?  

• Bei einer Barchart gibt es 2 Achse und deswegen auch 2 Attribute, aber in der Buch gibt es so ein Ausdruck, welche besagt, dass man für ein Barchart nur eine Attribute braucht. Was ist damit gemeint?  

• Until which point does it make sense to change the attribute of an object in a diagram to contain a message even to it loses some of its mathematical sense?  

• The author talks about efficiency in on chapter and how some channels are more effective than others. But Stevens show that the different aspects of channels could cloud the data. Doesn’t it make more sense to pick the channel depending on how detailed data has to be shown and then look for a more efficient channel? All together it doesn’t seem to be so good to talk about efficient when there are many more factors.  

• The author talks about some combinations of channel that work perfectly together and some that don’t work that well. But it seems to me that there is a huge potential of combinations with many diagrams. So my question is if the visualization science has even found all matches and if it would be possible to still find another perfect match to result in a new better visualization for a problem?  

• Can it be viable to use the same channel on the same mark multiple times for different attributes (e.g. the left side of a circle in a more or less saturated red and the other half in blue)? And if so, under which circumstances?  

• Using different channels with different effectiveness and popout properties help presenting the data in a way that corresponds to the underlying mental model (which was probably explored in the requirement analysis phase). But how can those inequalities amongst the channels be handled if the visualization is being made for data exploration, where there is no mental model yet and it is not known which attributes are more important than others?  

• If I have any other channel available, why would I choose motion for categorical attributes and how would I use it? 

Video

• Whats the speaker's name?  

• How does the light center/surround work?  

• Where there some notes of artists found, showing what was their theoretical knowledge they used for drawing? Or did they do all of these things heuristically?