I recently completed a fascinating usability study with the help of Dr. Stephan Weibelzahl and Sara Kyofuna at NCI in Dublin. Here is a report of our findings. NCI will be publishing an academic paper on the study.
Question: Are interactive infographics better than spreadsheets?
Study: In an eye-tracking study involving 32 participants, User Journeys (my company) and National College of Ireland compared two interactive infographics to their tabular data (spreadsheet) equivalents.
Conclusion: Contrary to popular belief, participants given the tabular data formats completed goal-oriented tasks faster and more efficiently than participants given the interactive infographics.
Overall, participants found the tabular format more usable, but felt the infographic format was more motivating and more satisfying.
Background
Interactive infographics are an increasingly popular way of displaying information online. For example, the New York Times newspaper has published over 200 interactive infographics. They are frequently used to visualise large data sets and make these more accessible for users. So how good are they really in getting across the facts?
I partnered with National College of Ireland to conduct an eye-tracking study that compared two interactive infographics with their tabular data format equivalents.
Method
The study involved 32 participants, 75% of whom were male.
We used the following infographics:
1. Snake Oil – An infographic depicting scientific evidence for popular health supplements
See: http://www.informationisbeautiful.net/play/snake-oil-supplements/
Snake Oil infographic used in study with kind permission of David McCandless.
Snake Oil – Tabular Format Equivalent
2. Baby Name Popularity – An infographic depicting the most popular baby names in Ireland during the years 1998 – 2010
Baby names infographic created using tutorial provided by Nathan Yau: http://flowingdata.com/2009/12/09/how-to-make-an-interactive-area-graph/
Baby Names – Tabular Format Equivalent
At the start of each test, we presented the participant with one of the infographics or with its corresponding tabular data format.
We then gave the participant a set of goal-oriented tasks, such as:
- “How many babies were named Aoife in 2008?”
- “How strong is the scientific evidence that vitamin D can reduce cancer?”
We also conducted a post-test survey, and asked participants to rate their experience in different categories.
Results
Overall, participants performed significantly better on goal-oriented tasks using the tabular format.
- Better performance with spreadsheet format: Participants who were given the tabular data responded with an average of 73% and 88% correct answers for the baby names and snake oil data. By contrast, participants who were given the infographic format averaged at only 49% and 63% correct answers respectively.
- Quicker with spreadsheet format: The average time for task completion in tabular format was 35.3 seconds and 43.6 seconds for baby names data and snake oil data respectively.
- Completing tasks with the infographic format took longer, averaging at 59.1 and 69.5 seconds.
Responding to the survey, users gave the tabular format a higher rating for the following categories:
- Usefulness e.g. it allows me to find information I want
- Ease of use e.g. it is easy to find what I am looking for
- Learnability e.g. it is easy to learn how to use this
However, participants gave the infographic format a slightly higher score in the following areas:
- Satisfaction e.g. I like this information presentation
- Motivation e.g. it encourages me to explore the data
Conclusion
While interactive infographics are visually attractive, they are not necessarily better at providing factual information. For simple goal-oriented tasks, our study indicates that the tabular format is more accurate, efficient and usable than the interactive infographic format. However, our study has also found that user satisfaction and motivation are higher with the interactive infographic format.
Further studies would be required to see how interactive infographics perform for discovery-oriented tasks.
This research was partly funded by Enterprise Ireland.