So, how do you define the minimum value?
Tuftes’ data density is really about three principles: (1) Above all else, show the data, (2) Maximize the data-ink ratio and (3) Erase non-data ink. In its extreme this could be interpreted as small as possible human could read. We’re talking about font-sizes as small as 3 pixels, but practically 5 pixels which is possible to read – looking like this:
In bright contrast stands accessibility. Not a lot of users would find it comfortable and effective to access information that way. It’s very hard to read and the risk of reading wrong increases with every pixel you lessen your font size. Thus, we should drive towards more effective communication of data than data-ink ratio.
User Experience is often the magic of combining different objectives into one design solution, achieving as different goals as accessibility, marketing, task completion, site objectives, revenue, conversion rates, user satisfaction and company design guidelines.
Have you ever seen a study where they have employed data density to compare different UI elements?
To find the most effective information visualization, we need to measure. Measuring in itself can be quite difficult… is the same information represented in Graph A better or worse than Graph B? Fortunately there is a study made in 2009: Measuring effectiveness of graph visualizations: a cognitive load perspective:
The construct of cognitive load in the context of graph visualization is proposed and discussed. A model of user task performance, mental effort and cognitive load is proposed thereafter to further reveal the interacting relations between these three concepts. A cognitive load measure called mental effort is introduced and this measure is further combined with traditional performance measures into a single multi-dimensional measure called visualization efficiency.
The study by Luzzardi and Del Sasso Freitas titled “An Extended Set of Ergonomic Criteria for Information Visualization Techniques” from 2004 come to a different conclusion:
Evaluating user interfaces is usually accomplished to detect design problems in the layout as well as during interaction. In information visualization techniques, interface usability issues are intertwined with the expressiveness of the visual representation. These two aspects are equally important and have to be evaluated in order to verify how well a visualization technique supports users’ tasks.
But neither of the studies say “yes” to your second question. There might be, but none as prominent and famous as it could make the first 50 hits of Google Scholar. I’d say you either have to make prototypes yourself and do a simple A/B-testing to find out what your users find best or do a real research at the university – maybe as a degree thesis. It would be interesting to read such a thesis.
