Some Guidance for Writing alt text for Data Plots

Krisanat Anukarnsakulchularp and Di Cook

Why?

Alt text provides a textual alternative to non-text content in HTML documents. It serves various purposes:

it is necessary for accessibility guidelines.
ensures your data visualisations communicate to everybody.
assistive technologies can convert alt text into other formats such as speech or Braille, providing a description for people using screen reading software.
alt text is displayed in place of the figure if it fails to load, or is viewed in a text-based browser.
alt text can assist both general and image-specific search engines.

Ingredients for writing alt text

The purpose of alt text is to describe what is in a plot for a person who cannot see it. Generally, this means very basic explanation of the elements of the plots. You wouldn’t typically interpret the plot, especially if it is also accompanied by a caption. A caption assumes that a person can see the plot, and provides additional context about why it is included and what is the most important thing to learn. Thus, the role of alt text is to provide the details of the plot elements for someone unable to see it, which would help to make the caption also more understandable. When a caption doesn’t exist, then the alt text has to play both roles.

The basic ingredients for alt text are:

What kind of visualization is it (line graph, scatter plot, bar chart…)?
What variables are on the axes?
What is the range of the data?
What is the aspect ratio.
How are variables mapped to elements such as points, lines, violins or boxplots?
What are the visible patterns (points are clustered in the top left and bottom right, lots of variability in size of the boxes of the boxplots, violins show a bimodal pattern, …) without interpreting them.

Examples

The examples are organised to show what would be written for some common plots, in both alt text and a caption. Below is one example, shown with different additions, and there are more examples available in the Examples page.

Static scatterplot

The plot is generated from this code.

library(ggplot2)
library(palmerpenguins)
p <- penguins |>  
  ggplot(aes(x = bill_depth_mm, y = bill_length_mm, 
             color = species)) + 
  geom_point() +
  scale_x_continuous(breaks = seq(13, 22, 2)) +
  scale_y_continuous(breaks = seq(30, 60, 10)) +
  theme(aspect.ratio = 1)
p

Alt text: A scatter plot of the penguin’s data from the palmerpenguin package. The plot has a bill depth (30 to 60mm) on the x-axis and a bill length (13 to 22 mm) on the y-axis. There are three species of penguins with different colour codes, Adelie (red), Chinstrap (green), and Gentoo (blue). The Gentoo penguins have an elliptical shape pointing to the top right, with the bill length ranging from 40 to 60 mm and bill depth ranging from 13 to 17.5 mm. The Adelie penguins have an elliptical shape pointing to the top right, with the bill length ranging from 15 to 22 mm and bill depth ranging from 30 to 48 mm. The Chinstrap has an elliptical shape pointing to the top right, with the bill length ranging from 40 to 60 mm and bill depth ranging from 16 to 22 mm.

Caption: Examining the differences in bill sizes of three species of penguins. Colour indicates species. Gentoo bills are smaller in depth but longer. Chinstrap and Adelie bills have similarly large depth but Chinstrap bills are longer.

There are more examples available in the Examples page.

Using BrailleR to partially auto-generate

The BrailleR package provides a text description of a plot that is generated using ggplot2. Here is the description generated for the plot shown above.

library(BrailleR)
VI(p)
## This is an untitled chart with no subtitle or caption.
## It has x-axis 'bill_depth_mm' with labels 13, 15, 17, 19 and 21.
## It has y-axis 'bill_length_mm' with labels 40, 50 and 60.
## There is a legend indicating colour is used to show species, with 3 levels:
## Adelie shown as strong reddish orange colour, 
## Chinstrap shown as vivid yellowish green colour and 
## Gentoo shown as brilliant blue colour.
## The chart is a set of 342 big solid circle points of which about 95% can be seen.

This is focused on how the plot is constructed, which matches half of the content. Note that, there are differences to the alt text we manually wrote. The missing part is the description of the patterns in the plot that can be seen. This is very difficult to automate. Thus the BrailleR package can provide a start to the alt text but additional information needs to be added.

Resources

There are many sites describing the purpose of writing alt text. None provides concrete explanations of documenting data plots. Here is our list of general resources related to alt text:

How to contribute

We welcome contributions! If you have a suggestion or addition please post this as an issue on GitHub.