The physical model on the left is a bivariate histogram showing the correlation between the heights of fathers (horizontal axis) and sons ("vertical" axis). This data was famously collected by Karl Pearson and Alice Lee between 1893 and 1898. The physical visualization is thought to have been constructed around this time period or soon after, possibly under the supervision of Pearson. It is kept at the Department of Statistical Science, University College London, founded by Pearson in 1911. I (Pierre) learned about this model after seeing a more recent model in the archives of the collection of solid mathematical models at the Institut Henri Poincaré (right image). A label said "Model Prepared by Dept. of Statistics, University College London". I then wrote to Professor Richard Chandler, current chair of the statistics department, to ask if he had more information about this model. He replied soon after: Thank you for your message below. This is interesting … I have never seen this particular model before, but the labelling looks very similar to some exhibits that were prepared for an event at the Royal Statistical Society in (I think) the 1980s. I arrived at UCL in 1994: at that time there were some people in the department who had been involved in that event. Unfortunately they are no longer with us, so I can’t provide any further information. I can, however, speculate that the model you’re showing is a cardboard copy of a much older one that was prepared at the end of the 19th century, probably under the supervision of Karl Pearson. We still have this. I found this particular model in my desk drawer when I first arrived at UCL: it was broken into a very large number of pieces, but I had it restored. It is quite a nice piece of 19th-century data visualisation, I think! Upon closer examination, the newer model does not show the same dataset as the older model. The following R code replicates the older model exactly, confirming that it shows the original dataset from Pearson and Lee: library(dplyr) library(HistData) library(plot3D) d = HistData::PearsonLee %>% filter(par == "Father" & chl == "Son") %>% reshape2::acast(parent~child, value.var="frequency") plot3D::hist3D(z = d[nrow(d):1,], col="white", border="black", shade=0.2, phi=60, theta=0) Sources: Pearson, K., & Lee, A. (1903). On the Laws of Inheritance in Man: I. Inheritance of Physical Characters. Left photo kindly sent by Richard Chandler. Right photo taken by Pierre Dragicevic during a visit kindly organized by Alba Málaga and guided by François Apéry.

American economist and statistician Karl G. Karsten patented a method for creating physical visualizations of temporal data by stacking two-dimensional plates, each representing a time period. Each of the plates shown above is a variable-width column chart representing the state of the stock market at the end of a particular month. Each bar is a type of stock, where the height of the bar encodes the stock price and its width encodes volumes of sales. Each month, a new layer is added. Seen from above, this 3D visualization becomes a 2D stacked bar chart showing volumes of sales across time. Karl G. Karsten wrote a book entitled Charts And Graphs in 1925, where he discusses physical visualizations used in engineering. Many of these were previously discussed in John Peddle's 1910 book The construction of graphical charts and Willard Cope Brinton's 1914 book Graphic Methods for Presenting Facts. Source: Karl G. Karsten (1926) Tridimensional Chart. US Patent. Related: Also see our entries 1935 – 3D Visualizations of Power Consumption and 1951 – Electricity Generated or Demanded.

[Data Beyond Vision] explores new ways of engaging with a dataset and the arguments and narratives behind it, in order to challenge the dominant paradigms of conventional screen-based data visualization. The project currently comprises: 3D printing a model of library member activity over time from the Shakespeare and Company Project juxtaposing documented activities from two sets of archival materials Folding paper forms of borrowing activity from the Shakespeare and Company Project surfacing the activity of women and and non-famous members Weaving representing intertextuality based on references in Jacques Derrida’s de la Grammatologie from Derrida’s Margins Images from left to right and top to bottom: (1) some of the data physicalizations designed in the course of the project; (2) lollipop chart of Shakespeare and Company membership; (3) unit origami volumetric representation of Shakespeare and Company membership; (4) folded model of Shakespeare and Company lending library membership activities; (5) weaving Derrida’s references. Sources: Koeser, Rebecca Sutton, Gissoo Doroudian, Nick Budak, and Xinyi Li. “Data Beyond Vision.” Startwords, no. 1 (October 2020). https://doi.org/10.5281/zenodo.3713670. Koeser, Rebecca Sutton, Nick Budak, Gissoo Doroudian, and Xinyi Li. “Data Beyond Vision,” July 11, 2019. https://doi.org/10.5281/zenodo.3261531.

A physical violin plot created by sculptor and psychology researcher Hunter Brown. See related entries also using clay here. Source: Tweet from Julia Strand (@juliafstrand), June 12, 2019.

Tattoos that Change Color When Reacting With Glucose Levels: minimally invasive, injectable dermal biosensors were developed for measuring pH, glucose, and albumin concentrations. Sources: Biotech (2019) Twitter post. Yetisen et al (2019) Dermal Tattoo Biosensors for Colorimetric Metabolite Detection.

An art installation in the Cathedral of Schwäbisch Gmünd (Germany) shows COVID-19 deaths as nails hammered into wooden cubes. More nails are added as the number of deaths increase. The text down the steps says "Fürchtet euch nicht", meaning "do not be afraid". Sources: Tweet from Friedrich Hart (@mxfh), Dec 9, 2020. SWR (2020) Schwäbisch Gmünd: 13.000 Nägel für Coronaopfer. Related: Also see our other entries on single-datum physical visualizations and on conveying deaths.