Experiments in Visualization

by Alison Hedley

In the summer of 2018, I used RAWgraphs to make a series of visualizations that accompanied a paper given for the Canadian Society of Digital Humanities (CSDH-SCHN). These mark the beginning of a new phase of the Personography, one that involves more in-depth visual analysis of our dataset.

Visualization of assigned sex frequencies in dataset of Y90s contributors. Turquoise is male; green is female; purple is unknown and red is other.

This pair of packed circle visualizations was created using RAWGraphs. The two images display the distribution of assigned sex across the persons in our data, which encompasses four categories: m (male), f (female), u (unknown), and o (other). The graphic on the left includes only the assigned sexes of magazine contributors; the graphic on the right includes the sexes of all magazine contributors and non-contributors who are in the Personography. As you can see, the distribution is almost the same in both cases, and the ratio is quite skewed toward males.

Visualization of assigned sex frequencies in dataset of all persons (contributors and relations). Turquoise is male; green is female; purple is unknown and red is other.

The skew of the data values is consistent with gendered expectations for late-nineteenth century authors, artists, and editors. The prevalence of the assigned sex of male also reflects the conventions of the historical record. In general, males were more visibly culturally and politically active than women at this time. Documentation of men’s activities was more valued and, in turn, more robustly archived. The visualization also tells us that our own Personography, drawing as it does on archives and biographical records, maintains this bias toward persons assigned male. Acknowledging this limitation of our data model is important to recognizing how it can and cannot help us understand fin-de-siècle literature and culture in more depth.

Visualization displaying every unique string of occupations that appears in the dataset of Y90s magazine contributors, organizing them by assigned sex.

This circular dendrogram was created using RAWGraphs. The visualization displays every unique string of occupations that appears in the dataset of contributors, organizing them by assigned sex. As you can see, the array of occupations held by females was much larger than the array held by males: female contributors to the Personography held a wider variety of occupations than did males. This pattern may be the result of gendered workforce expectations in 1890s Britain. It was easier for men to access the education and socio-professional networks needed to get and keep a vocational job, so women were more likely to have multiple occupations, even though a wider range of occupations was available to men to select from.