Survey results: A study of data visualization knowledge amongst network security professionals

The following survey was administered during the fall of 2011 with the goal of determining which visulization types are most and least preferred by professionals in the network security industry. Survey results are based upon familiarity, interpretability, and usefulness.

A second surveyed was administered to the general publlic with the same set of data visualizations as a comparison (these findings are located in the appendix of the slide deck).

Our Findings
Among all participants familiar with these visualizations:

• Pie charts, bar charts, line charts are most familiar types of visualizations
• Parallel coordinates, tree maps and sparklines are the least familiar types of visualizations
• Line charts are most useful, followed by bar charts, link graphs and histograms
• Bubble charts, parallel coordinates, tree maps and radar charts are least useful
• Radar charts are the most over-rated, followed by area charts, and pie charts*
• Parallel coordinates are the most under-rated, followed by heat maps and link graph*

* Based upon disparity between familiarity and usefulness
(For the complete report, view the Slideshare presentation below)

Our conclusion
Although many of the advanced data visualizations examined in this survey were ranked lower in terms of familiarity, interpretability and usefulness, we are not suggesting they should be avoided in daily use. Varied visualizations are instrumental when used to explore content in different contexts and  perspectives.

To maximize the effectiveness of less familiar data visualizations, users need to be educated on their applications. "Getting started" tutorials, inline help, tool tips and demos are just some of the materials that can be used to increase the data visualization user experience.

The opinions expressed here are my personal opinions.Content published here is not read or approved in advance by RSA and does not necessarily reflect the views and opinions of RSA nor does it constitute an official communication of RSA.

Data visualization short v1.1

View more presentations from Adam Winkler.

Using web analytics to undersand your users? Don't!

Using web analytics to predict user behavior is like a sending a jury to deliberate a trial without hearing any witnesses. Web analytics will give you the facts, such as the "who," the "when," the "what," and the "where," but it will never answer the "why."  To under stand user behavior - a.k.a the "why", you have to start with the users themselves. Usability tests, contextual inquiry, and interviews are always the best methods for user behavior analysis and is really the only way to understand motivation.

Web analytics, however, are a great compliment to these real user behavior studies, and create a solid basis when establishing a usability test plan. Understanding what users did on your site, where they went (or didn't go), when they did it, and who they were, will help pinpoint exactly what to focus on during real user studies.

The opinions expressed here are my personal opinions.Content published here is not read or approved in advance by RSA and does not necessarily reflect the views and opinions of RSA nor does it constitute an official communication of RSA.