Skip to main content

Critical Digital Literacies: Algorithmic Literacy

Reading for Algorithmic Literacy

Sources

Algorithm. (2018). In Encyclopædia Britannica

Ayre, L. B., & Craner, J. (2018). The Baked-In Bias of Algorithms. Collaborative Librarianship, 10(2), 76+

Badke, W. (2012). Personalization and Information Literacy. Online, 36(1), 47–49.

Carrington, V. (2018). The Changing Landscape of Literacies: Big Data and Algorithms. Digital Culture & Education, 10, 67–76.

What is an algorithm?

An algorithm is a systematic procedure that produces, in a finite number of steps, the answer to a question or the solution of a problem (Britannica Academic, 1999). 

In computer science, algorithms may provide search results, content recommendations, or targeted advertisements based on an individual user's past actions, as well as by grouping an individual user in with a representative cohort. Many algorithms draw on big data to identify patterns and make predictions about user behavior.

Algorithms may save a lot of time for users, or help them learn about new ideas or products they may not have otherwise explored. 

Algorithms can also reinforce existing biases, especially on social media. For example, clicking on an article that a friend posted on Facebook that leads to a liberal news source "teaches" the Facebook algorithm that you are interested in liberal news sources. In the future, Facebook may structure your news feed to include more liberal news sources and filter out conservative news sources. This can lead to a "filter bubble," in which users are only presented information that confirms their existing beliefs (Ayre & Craner, 2018). 

When the Chief Technology Officer of Reddit wanted to illustrate the complexity of his platform's algorithm in 2017, he used this GIF of a virtual Rube Goldberg machine:

(Image via Reddit)

What is algorithmic literacy?

Algorithmic literacy involves:

  • Recognizing the inherent biases in computer programming
  • Critically evaluating the information we receive online, and not assuming that the highest-ranked information is the "best" information
  • Understanding that engaging with digital platforms involves sacrificing a degree of privacy

Most algorithms are a "black box," meaning that they are not open to scrutiny from users and are considered the property of the company that develops them. While users see the output of an algorithm (i.e., search results), they are not granted access to the input (i.e., the data that has been collected about them so far) nor the decision-making process that filtered the output.

Opting out of personalization algorithms

Search engines and social media platforms may use personalization algorithms to rank search results and decide which advertisements to show you, and may also determine which posts to show on your feed. Personalization is based on your prior searches and other internet browsing information collected via cookies.

To opt out of personalized ads on Facebook, follow the steps in this video:



There are also a number of browser plug-ins for Google Chrome and Firefox that you can use to opt out of personalized ads. Google Chrome also allows you to opt out of ads more simply than Facebook.

Algorithmic Literacy Resources

Algorithms and Bias