Algorithmic Literacy

Basic knowledge of the technical and socioeconomic processes behind algorithmic platforms provides clarity on how and why we see the information we do. And, equally important is examining the personal side of these processes – how our own cognitive, affective, and behavioral processes are impacted by algorithmically-mediated information. We suggest four algorithmic awareness competencies (Augustinus, 2022; Long & Magerko, 2020) as particularly important to develop ethical, effective processes for learning with AI:  

1. Identification of algorithms in daily information interactions 

2. Explanation of how algorithmic platforms function 

3. Evaluation of the strengths and limitations of algorithms 

4. Integration of ethical and social impacts in our use and regulation of algorithms  

These foundational skills and knowledge about the nature and functionality of algorithmic technologies can be approached broadly, in the context of personal information use, or in the context of how one interacts with information in a particular discipline or field. By examining 1) the processes by which information gets to us, 2) our own learning processes in relation to that information, and 3) the processes by which people participate in knowledge creation, the focus of learning shifts to developing a holistic understanding of the content, rather than merely providing the “right” answer to get the desired grade. A process orientation positions students to both demonstrate their learning and apply it across contexts. When learning how to learn is paramount, using AI tools to do the work for you can be less appealing. As students document the steps they take and the tools they use, receive feedback, work collaboratively, and improve, the process of learning becomes just as valuable as the product or answer itself. Confidence in the process is what equips students to be successful across contexts – academic, professional, or personal. 

Example activities to consider 

• Together as a class, explore or test out an AI tool’s capabilities and report out on findings or challenges encountered. At each step in the process, ask students to describe what they are doing and seeing, and why they think it might be happening. Discuss potential benefits or concerns of using the tool. 
• Explore differing outcomes of completing the same task if each student uses the same prompt with an AI tool or, alternatively, if each student writes a unique prompt to use the same AI tool. Explore why outcomes might be different for different users or different prompts. 
• Shared responsibility: As you consider AI use in your course, expand beyond the prescriptive “use AI for this; don’t use it for that” to a conversation and a question: “Why does it matter where AI comes into our work? Where is human decision making most important?”  

Reflect and apply: Process orientation 

• Where am I placing emphasis on demonstrating and evaluating the process of finding information or solving a problem?  
• Where am I placing emphasis on providing an answer or product?  
• How might I provide more opportunities for feedback, revision, and collaboration? 
• In my teaching context, at what point(s) is it most critical for students to differentiate and make informed decisions about the pros/cons of using AI? How might I guide them to explore the benefits of taking time to work through a process that might be easier or faster using AI? 

Zoe Mayhook, Assistant Professor, Libraries & School of Information Studies

Prof. Zoe Mayhook (Assistant Professor, Libraries & School of Information Studies) partnered with a student to develop a process-oriented workshop about empowering the critical evaluation and use of AI tools for research: “Working with a student partner made me appreciate that students are also thinking critically about the benefits and shortcomings of AI tools, including the limitations of using these tools as information sources. My student partner and I assessed various AI tools tailored to support academic research, which subsequently informed a workshop session on utilizing AI tools for literature reviews. Aligning with the algorithmic literacy process orientation allowed us to cultivate algorithmic awareness through the exploration of these AI tools, and then we engaged in shared decision-making in considering how students might best learn and benefit from this information.” Mayhook explained that the algorithmic awareness competencies ”equipped me with valuable approaches to engaging students in the evaluation and ethical use of AI tools, both in academic and professional contexts.”

Reflective practice benefits learning, professional work, and personal growth in many ways. It deepens and documents learning, provides clarity on the purpose for our actions, and informs our next steps (Ash & Clayton, 2009). This is particularly critical in algorithm-mediated information interactions, where the tech tries to remove as many steps of the information gathering process as possible. Algorithmic platforms tend to prioritize information that will generate an immediate, emotion-driven reaction in the audience. For example, social media feeds elevate posts that have garnered the most interactions (likes, retweets, etc). We can integrate reflective exercises into assignments or discussions that help form a habit of pausing to examine how our own preconceptions and emotions impact our perception of the accuracy or trustworthiness of information (Oeldorf-Hirsch & Neubaum, 2021). Reflection can look many ways: exit “tickets” to capture learning at the end of class; guided reading questions during a homework assignment; identifying one’s own previous experiences that inform the way one learns; peer- and self-evaluation; and more. Reflection serves multiple purposes, both enabling personal growth, creating a community of learners, and serving as formative assessment. This reflective approach to information enables us to retain our decision-making power amidst algorithmic systems that are designed to make decisions for us. 

Example activities to consider: 

• Provide structured guidance and expectations for students to engage in self evaluation. This could look like asking students to set goals for the semester and then inviting them to be part of evaluating how well they achieved those goals. When students’ goals are clear and personally meaningful, it may help when making choices about when or when not to use AI for coursework. 
• As is feasible, regular student check-ins with an instructor or TA will provide opportunities to identify areas of both strength and struggle, as well as build relationships. The benefits of these reflective meetings demonstrate that learning and information are socially networked, and that AI tools cannot replace the unique affordances of human-to-human interaction, but rather can supplement them. 
• When students engage in reflection, help them to move beyond simply reporting events or commenting with their opinion – provide guiding prompts that aid in identifying challenges and next steps in overcoming those challenges or building on their experience. When learning with AI, it is particularly important to develop the reflective practice of identifying “now what” – how will I apply and transform the information I’ve found for my particular context? AI tools make it very easy for our brains to “check out” but it is essential that we train our brains to view these tools not as decision-makers, but option-presenters. 

Reflect and apply: Reflection 

• When might I add moments for students to pause and notice, document, or share their own thoughts, actions, reactions?  
• How do my own emotions impact the information I decide to share with my students?  
• How can I model my own process of reflection and how it impacts my research and teaching? 
• In my teaching context, at what point(s) is it most critical for students to examine their motivations, assumptions, and previous experiences and how those influence their reasons for doing things? How might I help students identify a plan of action for using AI critically, taking these insights into consideration? 

Annaelle Gackiere (Psychological Sciences and Anthropological Sciences) says, “Reflecting on courses I have taken in the past, I remember noticing strong reactions against AI use in the classroom, to the point that I was not made aware of its complexities, the ethics tied to it, and the possibilities of its use in writing-heavy contexts.” She explains that partnering with a professor who was open to discussing the possibilities of using AI helped her enact the algorithmic literacy practice of reflection: ”This partnership changed my outlook quite drastically. Now, I look over the syllabus more critically and think about the reasoning behind the extent to which we can use AI. I also assess my behaviors and choices, thinking more deeply about what I would want to change about decisions made in my classes.”

The overarching goal of algorithmic literacy is to cultivate education for democracy – to prepare students to meet the world’s complex questions and varied perspectives with curiosity, open-mindedness, and a belief that their actions make a difference. Algorithmic literacy approaches create opportunities for ‘transformative action’ – collaborative, participatory learning experiences with real-world connections, which lead to taking action that challenges or improves the status quo. Students, like all of us, want to contribute their strengths and be challenged to do work that matters to others. Transformative action builds on the approaches mentioned earlier to create a learning environment where students view their course work as a vehicle to actively improve the world around them with their own ideas. Developing a process orientation in your classrooms develops the skills to put information in context, which is key to identifying how that information can be used reliably. Developing inquiry-based learning in your classroom develops a question-posing mindset toward learning with algorithmically-mediated information. These approaches bridge in-class and out-of-class information practices by inviting students to apply prior knowledge they bring from unique experiences to their work in class. Students use algorithmic social media platforms all day, every day, and they have already developed skills and strategies for evaluating information and communicating ideas tailored to a particular audience (Head, Fister & MacMillan 2020). Some approaches to integrating transformative action could include: ask students to create examples or documentation that will support future students in the use of AI for learning; replace a final exam with an experiential or service learning project to create an AI tool that benefits a community; engage in a negotiated curriculum, where your students help determine when and why AI can be used, and can even sign up to teach a lesson or lead discussion around the use of AI. There are as many ways to make learning transformative as there are students – asking them what matters most to them is a great place to start. When our teaching and learning culminates in taking pro-social action to improve the world around us, it requires both instructors and students to draw on unique personal experiences and strategies for “real world” information evaluation. It creates opportunities for collaboration and learning from one another as we encounter new types of information and practice new skills.  

Example activities to consider 

• Rather than making a unilateral or inflexible decision about when and why to use AI in your course, collaboratively develop your course AI policy with your students. This can create an opportunity for students to take ownership of their own learning outcomes as they consider how AI might positively or negatively affect their ability to learn what they want to learn.  
• Consider ways that you might feature student work as examples, documentation, or study materials for future students. For example, if students document the prompts they used with an AI tool and discuss what was successful or unsuccessful, those experiences could be used as guidance for other students learning to use AI productively. 
• When developing assignments or projects, challenge students to take on applied, real-world problems or scenarios. Projects with meaningful outcomes for a broader community or for the students’ future work can be a huge motivator – importantly, this could include working with your students to identify what exactly they might find to be a meaningful outcome, how they envision AI as a support mechanism achieving it, or if AI might be a hindrance in some cases.  

Reflect and apply: Transformative action  

• What is the most essential purpose or goal of the work I am asking students to do?  

• Is the purpose made explicit, especially if it could be perceived as “busy work” or some other undesirable task?  
• How might I provide opportunities for them to envision how their knowledge and experience can uniquely contribute to the work they are doing? 
• In my teaching context, at what point(s) is it most critical for students to influence the conversation around how to approach AI for learning? How might I involve students in designing ways of working with AI with outcomes that they find meaningful?