Infographic Analysis — an Information Vizualization Study

Overview

The examples chosen below had Few's classifications (time, ranking, part-to-whole, deviation, distribution, correlation, geospatial, and nominal) and Kosslyn's goals (connect with your audience, direct and hold the reader’s attention, and promote understanding and memory) applied to identify what the author wanted the reader to take away from the graphic and whatever the relationship(s) the author presented were a good choice for presenting the data.

Methods

Good Graph

In 2015, two shark attacks off the coast of North Carolina left two injured teenagers. In response to these attacks, ABC NEWS authors an article, Tips to Survive a Shark Attack to calm the fears of ocean loving patrons. In this article, the news organization developed a graph to show its readers shark attacks from 1900 through 2009 (Figure 01). They used recorded data for this graph through incident reports collected by the Florida Museum of Natural History at the University of Florida.

The author creates a relationship of activity to the total number of shark attacks over the duration of time. These relationships will be identified by Few’s classifications of time, ranking, and distribution. The classification of time is presented by the number of shark attack reports recorded from 1900 to 2009. The Y-axis presents attacks in chronological order and groups them by decade. The classification of ranking can be seen through the increased occurrence of events on surface recreationists. The victim’s activities will be counted and grouped categorically. By ranking these activities, the audience sees that before such flotation devices become popular (1980s), swimmers and bathers are more likely to be attacked. Finally, the classification of distribution will describe the victim’s activities within each decade. The authors had used horizontal bar charts within each decade to represent the distribution of victim activities.

The dependent variables within the graph presents an interval variable based on the date of a shark attack, while the second nominal variable categorizes the type of activity the victim had performed at the time of the incident. The independent variable, is an interval value based on the total number of victim activities in a given decade.

The author’s choice to use a bar graph is a solid choice. It shows the reader a comparison between the categories of time (shark attacks from 1900 to 2009), grouped attack types (surface recreationists, divers, swimmers & bathers, and victims upon entering or exiting the water) and total number of shark attack groups within a decade.

Discussions

Connecting with the Audience

The intended audience loves to recreate at the ocean, but may fear it based on recent shark attacks. The conveyed message, according to historical data, tells those who recreate on the surface (anyone who uses any type of floatation device) that they will be more vulnerable to attacks. The grouped categories help describe activities like surface recreationists, divers, swimmers & bathers, and victims upon entering or exiting the water with proper relevance and knowledge to the readers. This includes categorically linking past and present surface activities through an entire millennium. This graph upholds both principles of relevance and appropriate knowledge. The authors continue to communicate the right amount of information without overloading or constraining their readers.

Direct and Hold Attention

The authors allow their audience to find striking differences in the activities by using distinctive elements of design in the graph. The authors use the principle of salience through their choice of primary colors to depict each activity category during the attack. Their color selection helps the audience identify and follow activities through the course of time. All text, both on the X and Y axis, have been distinct in color, size, and font. These elements make identifying and comparing variables easier.

The authors upheld the principle of perceptual organization by using distinct aspects. Their use of zebra striping on the graph’s background gives better readability to decade groupings, thereby utilizing the grouping law of similarity. They include white strokes around each activity type, giving greater detectability between activity types within a decade. Their readers have easily identified specific types of activities within a decade by using the grouping law of proximity. Finally, for the reader to remember grouped elements, each of the activity types is presented in the same order. Each decade allows the use of perceptual organization of information from one decade to another. To reinforce an activities organization, the grouping of common fate is used to label and color each corresponding activity in the same logical order.  This graph did not contain any erroneous elements that could be considered “chart junk”.

Promoting and Understanding Memory

The authors uphold the principle of compatibility by their reader's ability to understand and recall surface activity attacks through time related events. Therefore the bar graph is an appropriate method. A principle of compatibility is demonstrated through the use of more is more. The authors show total numbers on the X-axis combined with the length of the bar to show increasing amounts of shark attacks on surface activities through the decades and not just the past 10 or 20 years.

The framework and content of the graph can easily be deciphered and inferred without labels. However the presence of labels will strengthen the overall informational structure.

Recommendations

• Add labels to the X and Y axis
• Redesign the legend to be little less prominent
• Add more space between the decades to show better decade distinction
• Simplify the Y-axis from reading the decades as 1900 - 09 to something more simple (e.g. 1900s, 1910s, etc)

Bad Graphic

In 2015, a local municipal government held four long-range planning sessions to gather information to assist in the development of its Long Range Financial Plan for 2015-2017. The main objective of this document is to set forth long term goals, solutions, and strategies to address key issues that are used as guidelines during the annual budget process.

The Long Range Financial Plan describes in one section how the majority of the government's expenses are salary and benefit related costs. The government authors have created a graph for their audience, the taxpayers, which illustrates historical General Fund expenditures by account classifications. The numbers were obtained through data tables found in annual budget documents.

The authors created three types of classification relationships which can be identified in this graph. They use (1) a classification of distribution to show total expenses from a range of seven fiscal years. (2) A classification of part-to-whole which highlights each account classification within the fiscal year’s total expenditures. This stacked-bar chart demonstrates total expenditures within a given fiscal year. The chart also categorizes the total expenses into seven, colored categories with included values. And (3) A classification of time demonstrates how the total expenses increase from one year to the next. The visual data allows the reader to follow the upward trend of expenses each fiscal year.

The dependent variables in the graph present (1) an interval variable based on the year of the government's budget and (2) a nominal (showing a category name) and ratio (label value in millions of dollars) variable which categorizes the type of expenditures that comprise an entire year’s fiscal budget with values. The independent variable, is an interval value based on the total number of expenditures in a given fiscal year.

The author’s choice in using a stacked-bar graph is good, considering the trend analysis and account classifications it illustrates when showing the total cost of expenses. The reader can easily identify and compare categories of time (fiscal years from 2009 to 2014), account classifications (salaries, health, pension, fleet maintenance, equipment, risk, and all other) and total expenditures per year.

Discussions

Connecting with the Audience

The government's authors upheld a small portion of the principle of relevance. They explain through seven account classifications (salaries, health, pension, fleet maintenance, equipment, risk, and all other) how overall expenditures break down. These categories convey relevant knowledge to the reader in understanding the total General Fund expenses for each fiscal year. However, even with this knowledge having been conveyed to taxpayers that personnel costs will likely increase due to costs of salaries, public pensions, and health benefits; the authors violate the principle of relevance by giving taxpayers too much visual information and overwhelming them with account classification values.

Direct and Hold Attention

The element of colors may have upheld the principle of salience by drawing attention to the stacked-bars, but all legends, axis, and text values violate this principle with no aspects strong enough to call out important information in the chart. The colors, referencing account classification in the horizontal legend, have no significant relationship to their fiscal year breakdown. Since the taxpayer is their audience, the background color selection and foreground content for these categories may be difficult for the visually impaired.

After researching accessibility websites like WebAIM and W3C, they recommend contrast ratios of 4.5:1 for normal text at 12 point. The chart had been measured for background and foreground contrast ratios. The results had ranged from 19.75:1 to 2.4:1. With three category colors falling below recommended contrast levels and all font sizes below the normal text size; this clearly reduces the audience's ability to read and comprehend the makeup of total expenses and therefore violates detectability as well as the principle of discriminability.

Promoting and Understanding Memory

The violation of the principle of capacity limitations has been seen by the placement of legend and the value of account classifications. The reader requires additional information to understand the category colors in order to decipher the graph. The legend’s scattering location and multiple classifications makes it difficult to create correlations between the chart and legend. These actions violate the principle of capacity limitations by using the aspect of short term memory limits.

The framework and its measureable elements are easily decipherable, but could use additional labels to better define the independent and dependent variables. The content’s background may require additional horizontal markers or inner grid lines. Since the author’s time, categories, and trends include multiple classification, their audience may find additional markings helpful in deciphering yearly classification comparisons.

Recommendations

• Add labels to the X and Y axis
• Increase font size on X and Y axis
• Increase font size for chart name label
• Increase font size of total expenses value
• Redesign the legend to be less prominent, no background or border
• Redesign legend to flow with graph for reader’s knowledge of categories
• Reposition the legend to limit the reader’s mental model
• Recolor legend categories, sort in color order that represents graph for easier reference
• Remove unnecessary classification values and place in referenced data table below graph. This allows all types of audiences, with or without impairment to reference values
• Show total expenses more prominent on top of stacked-bar chart to define total for year
• Close the space between fiscal years to show better year over year comparisons