Histogram binning revisited with a focus on human perception



Raphael Sahann, Torsten Möller, Johanna Schmidt

 External link (DOI) 

 View presentation:2021-10-28T15:50:00ZGMT-0600Change your timezone on the schedule page
2021-10-28T15:50:00Z

Exemplar figure, described by caption below — Revisiting histograms with a focus on human perception. We evaluated how well human viewers can detect the underlying data distribution in a histogram when different sample sizes and bins are used. For this, we created datasets with a different number of samples (first row: few, last row: many) and a different number of bins (left column: 2, right column: 100). In this Figure a bimodal distribution was used to create the datasets in this illustration. We could show that most sample-based algorithms overestimate the number of bins which are necessary for human viewers.

Keywords

Perception & Cognition, Guidelines, Human-Subjects Quantitative Studies

Abstract

This paper presents a quantitative user study to evaluate how well users can visually perceive the underlying data distribution from a histogram representation. We used histograms with different sample and bin sizes and four different distributions (uniform, normal, bimodal, and gamma). The study results confirm that, in general, more bins correlate with fewer errors by the viewers. However, upon a certain number of bins, the error rate cannot be improved by adding more bins. By comparing our study results with the outcomes of existing mathematical models for histogram binning (e.g., Sturges’ formula, Scott’s normal reference rule, the Rice Rule, or Freedman–Diaconis’ choice), we can see that most of them overestimate the number of bins necessary to make the distribution visible to a human viewer.