Teaching:TUW - UE InfoVis WS 2010/11 - Gruppe 01 - Aufgabe 2

Graphical Inference for Infovis

The following article summarizes the work of [Wickham et al., 2010] on graphical inference.

Introduction

Information visualization provides tools to show new relations in data. Statistics instead provides methods that can examine if an assumption is correct or not. Graphical inference tries to find a balance between these two methods. With the help of apophenia, the capability of human to detect patterns in noise, hypotheses can be established. The goal of graphical inference is, as in statistics, to reveal faulty conclusions.

Motivation

Statistical methods generally try to show that a hypothesis is true or not. More specifically statistic investigate whether a difference exists (testing) or how big the difference is (estimating). For graphical inference you want to know whether a difference is actually here, so graphical inference works as testing procedure. In statistics this is called a null hypothesis H0 (the situation) and the alternative hypothesis (the assumption). The result of a statistical test can take two fault conditions:

• a H0 is rejected, although a H1 is not true (also called type I error)
• a H0 is no rejected, although a H1 is true (also called type II error)

The testing process in statistics can be compared with the criminal justice system, where an accused is judged guilty or innocent. During the trial the defense tries to show that the null hypothesis is true, the prosecution advocates the alternative hypothesis.

The static test compares the accused and known innocents, using a specific metric. To assess the guilt of the accused, the ration fo the innocent that look more guilty than the accused is computed. A type I error would be a convicted innocent and a type II error would be an acquitted guilty.