Implicit Error, Uncertainty and Confidence in Visualization: An Archaeological Case Study.

While we know that the visualization of quantifiable uncertainty impacts the confidence in insights, little is known about whether the same is true for uncertainty that originates from aspects so inherent to the data that they can only be accounted for qualitatively. Being embedded within an archaeological project, we realized how assessing such qualitative uncertainty is crucial in gaining a holistic and accurate understanding of regional spatio-temporal patterns of human settlements over millennia. We therefore investigated the impact of visualizing qualitative implicit errors on the sense-making process via a probe that deliberately represented three distinct implicit errors, i.e., differing collection methods, subjectivity of data interpretations and assumptions on temporal continuity. By analyzing the interactions of 14 archaeologists with different levels of domain expertise, we discovered that novices became more actively aware of typically overlooked data issues and domain experts became more confident of the visualization itself. We observed how participants quoted social factors to alleviate some uncertainty, while in order to minimize it they requested additional contextual breadth or depth of the data. While our visualization did not alleviate all uncertainty, we recognized how it sparked reflective meta-insights regarding methodological directions of the data. We believe our findings inform future visualizations on how to handle the complexity of implicit errors for a range of user typologies and for highly data-critical application domains such as the digital humanities.

dendsort: modular leaf ordering methods for dendrogram representations in R.

Dendrograms are graphical representations of binary tree structures resulting from agglomerative hierarchical clustering. In Life Science, a cluster heat map is a widely accepted visualization technique that utilizes the leaf order of a dendrogram to reorder the rows and columns of the data table. The derived linear order is more meaningful than a random order, because it groups similar items together. However, two consecutive items can be quite dissimilar despite proximity in the order. In addition, there are 2 (n-1) possible orderings given n input elements as the orientation of clusters at each merge can be flipped without affecting the hierarchical structure. We present two modular leaf ordering methods to encode both the monotonic order in which clusters are merged and the nested cluster relationships more faithfully in the resulting dendrogram structure. We compare dendrogram and cluster heat map visualizations created using our heuristics to the default heuristic in R and seriation-based leaf ordering methods. We find that our methods lead to a dendrogram structure with global patterns that are easier to interpret, more legible given a limited display space, and more insightful for some cases. The implementation of methods is available as an R package, named "dendsort", from the CRAN package repository. Further examples, documentations, and the source code are available at [https://bitbucket.org/biovizleuven/dendsort/].