- Mar 2025
-
www.digitalhumanities.org www.digitalhumanities.org
-
test broad literary-historical hypotheses in a way that resists confirmation bias
Looking through the article, I've began to wonder something that it does not entirely explained or something I've missed: how does one test hypothesis/experiment when you're dealing with abstract items (i.e. the narrative/text of literature, not the medium) that can't really be changed? Sure you could be able to record something such as changes over time in the ideas or methods used by authors, but that would be closer to an observational study such as a survey.
-
I have chosen distant reading because the phrase underlines the macroscopic scale of recent literary-historical experiments, without narrowly specifying theoretical presuppositions, methods, or objects of analysis.
I'm not going to lie, while I see how distant reading could be seen as connected to digital humanities, especially since it seems to focus on the general idea of studying trends in literature of the same era, I'm not sure how people made the over-association in the first place. Personally, it still seems possible to do distant reading without needing digital tools (such as text mining).
-
- Feb 2025
-
giorgialupi.com giorgialupi.com
-
Data represents real life. It is a snapshot of the world, in the same way that a picture catches a small moment in time.
I hold a pretty similar sentiment towards social history, with namely in holding value in the daily lives of other and similar cases, so I appreciate how this is reflected in related disciplines.
-
Let’s just stop thinking data is perfect. It’s not. Data is primarily human-made. “Data-driven” doesn’t mean “unmistakably true,” and it never did.
I would say that's not only the cause of human error, but possibly even the unusual cases that lie outside the "big net" of a large sample size, assuming the data set in question was lucky enough to have one.
-
Making enticingly accurate infographics requires more than a computer drafting pro- gram or cut-and-paste template, the art of information display is every bit as artful as any other type of design or illustration
I definitely understand why so many could easily fall into the trap of defaulting to a simple excel chart or the like, although I do think the graphic creator should be given such context behind the data before hand as a means to help the avoid the trap.
-
-
www.datavis.ca www.datavis.ca
-
We recently discovered a remarkable and little known graph portraying the deaths among passengers and crew on the maiden (and only) voyage of the RMS Titanic. This was published in the London illustrated newspaper, The Sphere, on May 4, 1912, less than one month after the ship sank. PDF of Sphere graph
Truth be told, I honestly consider this graph as one of my favorites from this page. While it doesn't incorporate elements such as time or positioning, it presents the data in a reasonably clear fashion.
-
Parallel sets Parallel coordinate plots provide a way to display multidimensional data in 2D plots. They do this by representing the variables as a set of parallel axes, and showing each observation as a line in parallel coordinate space, rather than as a point in standard coordinate space. Extensions of this idea for categorical data led to “parallel sets plots”, and some variations, a number of which use the Titanic data for examples. Bendix, Kosara, and Hauser (2005) Parallel sets: Visual analysis of categorical data and Kosara:2006-parallel Parallel sets: Interactive exploration and visual analysis of categorical data developed an interactive system to explore multivariate categorical data using parallel sets, in which the lines between categories of successive variables are of width proportional to the joint frequencies.
Due to the lack of visual clarity, I struggled to understand what 2005 parallels sets were actually representing in this context (especially when external searching seems to tell me that these types of plots are usually formatted horizontally), to the point of forgetting how most of these charts are tracking how of a certain grouping lived/died from the sinking, which makes me question on what benefits we get from them. I do appreciate the 2013 charts not only for an accurate line widths, but being clear enough with the color and shade distinctions in certain lines to make clear what feeds into what (although I do wish the "Survived" category was either on top or bottom rather than the middle).
-