Source: Effectiveness of Feature-Driven Storytelling in 3D Time-Varying Data Visualization
The authors designed an experiment to evaluate the performance of feature-driven animation and interactive systems, in order to compare the accuracy and time required to reach results.
Their results suggest that feature-driven storytelling animations may consistently lead to more timely results, and that they have comparable accuracy to interactive visualization across a variety of tasks common to scientific visualization.
According to the guidelines used for creating the animations:
- Start with an overview of entire duration
- End with a focused view in relevant time duration
- Include all relevant data attributes in at least one
segment of the animation sequence
I can see why it’s not really a surprise that the use of animations could lead to more timely results, as the guidelines require a domain expert to previously analyze which data is considered relevant and how should it be approached during the animation, but it also leads me to believe that animation based analysis can be more limited than an interactive system, and even accentuate the possibility of errors if they are not correctly done.
I was wondering what your opinion is on this? And also, do you know of any modern guideline/technique for generating feature-driven animations, preferably being used in a working system?