# Excutive Summary of Chapter: Graphical Integrity

By: Ahmed Abubakr

# Introduction

Graphical integrity chapter aims to help the designer to create clear and efficient graphs by using important principles and tools.

# Distortion in a Data Graphics

Distorted graph is the graph that shows inaccurate visual representation of numerical quantities.
There are two goals that need to be achieved when creating graphs which are:

1. The representation of numbers as physical objects on graphs should be proportional
to the numerical quantities that are being represented.
2. Labels on the graphs should be clear and detailed with explanation.
Misrepresentation of the numerical data on a graph can be measured by the Lie Factor formula.
Dividing the size of effect shown in graphics by the size of effect in data would results the lie factor.
If the lie factor is equal to one then the graph accurately representing the numerical data, however
if the lie factor is greater than 1.05 or less than 0.95 this might indicate in distortion.

# Design and Data Variation

Each part in the graph generates expectations about the other parts and and that determine what the eyes see. Confounding of
design variation with data variation results in deception, therefore the focus should always be on
showing the data variation NOT design variation.

# Visual Area and Numerical Measure

Ambiguity can happen when using areas to show one-dimensional data. That kind of designs should be
avoided by making sure that the number of variable dimensions is not more than the number of dimensions used in the data.

# Context is Essential for Graphical Integrity

For clarity, data being represented in the graph should tell the reader what the data is being compared
to, without any unnecessary data. The principle says:
Graphics must not quote data out of context

# Conclusion

Graphical integrity can be defined by six principles:

1. The representation of numbers as physical objects on graphs should be proportional to the numerical quantities that are being represented.
2. Labels on the graphs should be clear and detailed with explanation.
3. Show data variation, not design variation.
4. In time-series displays of money, deflated and standardized units of monetary measurement are always better than nominal units.
5. The number of information-carrying (variable) dimensions should not be more than the number of the dimension in the data.
6. Graphics must not quote data out of context.