October 31st, 4pm, the Thesis Defense: "Animated Transitions to Support Visualization of Missing Data" by Lea Tschiersch will take place at HCC Lab.

Abstract

Large datasets often contain missing data, which needs to be dealt with when visualizing the data, either by deletion or by depicting missing data in a certain way. These visualizations are often interactive. Animated transitions can be implemented to smoothen the switch between views.

In the context of the IKON project, which goal it is to facilitate knowledge transfer, animated transitions and missing data visualizations for a line chart will get conceptualized, implemented and evaluated. The general goal of the thesis is to enhance the way how users work with visualizations consisting missing data. Additionally, more information is tried be to collected regarding the user’s interaction with the chosen missing data visualizations.

This thesis explores the theory behind missing data, why data can be missing and what can be done to handle missing, including imputation methods and visualization techniques. It also examines already existing visualization techniques and studies, presenting their findings regarding preference of visualization types and different criteria such as accuracy and confidence in data.

Next, this thesis delves into the theory of animated transitions for data graphics. This includes the different types of transitions in data graphics and rules which should be followed to create successful animated transitions. Two different studies about animated transitions are be presented and their results are summarized, showing the possible positive effect animations can have for transitions.

Building on the theory of missing data and animated transitions in addition to the presented existing studies and the goal of the thesis, design requirements are set up. In the implementation “dashing” is chosen as missing data visualization and slow and simple-looking transitions are added to the line chart of the IKON project.

For evaluation a pilot study is conducted in which three participants interact with the implementation. It finds that all participants prefer having an animated transition over a static one. They were able to discern missing data from existing data. The pilot study, however, did also suggest that confidence in the visualization with missing data was high since participants were not hesitant to trust the missing data visualization.

With the pilot study conducted and an implementation in place this thesis presents a first glimpse on how animated transitions for missing data visualizations can be implemented, serving as a possible foundation for further studies.

The defence will be held in English.

First assesor: Prof. Dr. Claudia Müller-Birn

Second assesor: Prof. Dr.-Ing. Volker Roth

Location: HCC Lab, Königin-Luise-Straße 24/26, room 120