Supervisor: Torsten Möller
With Anscombe's Quartet [1] it was demonstrated quite figuratively that summary statistics can be very misleading or, at least, hard to interpret. Just recently, this example has become quite playful with the Dinozaur Dozen [2]. However, there are a number of statistical measures, that don't have an easy (visual) explanation. One of them is Krippendorf's alpha [3], a very common measure in the social science for measuring the agreement between subjective coders (as in labeling text or documents). The challenge of this project will be to:
- understand the measure
- develop simple alternatives
- develop different visual representations that "bring this measure to life", i.e. make it easy(er) to understand
Prerequisites: VIS
Contact: Torsten Möller
[1] Anscombe, F. J. (1973). "Graphs in Statistical Analysis". American Statistician. 27 (1): 17–21. doi:10.1080/00031305.1973.10478966. JSTOR 2682899, see also Anscombe's quartet.
[2] Justin Matejka, George Fitzmaurice (2017), "Same Stats, Different Graphs: Generating Datasets with Varied Appearance and Identical Statistics through Simulated Annealing," ACM SIGCHI Conference on Human Factors in Computing Systems. see also Same Stats, Different Graphs.
[3] Krippendorff, Klaus (1970). Estimating the reliability, systematic error, and random error of interval data. Educational and Psychological Measurement, 30 (1), 61–70. see also Krippendorff's alpha.