Microlearning: Transform content into bite-sized units with AI

Much has been written about the many distractors students face, including smartphones, tablets, and laptops that are ubiquitous in university classrooms and readily available to the online learner. Numerous studies provide evidence that when learners are distracted, it can have a detrimental effect on their learning [1, 2]. According to Schmidt, when students become distracted during learning, their brain processes and stores information in inefficient ways, “resulting in knowledge that is much less adept at extending and extrapolating to new contexts (decreased transfer)” [3].

In addition to the distractions students attend to, we must be mindful of how much information students can process during a class session or online learning module. If we provide too much content, there is a risk that students cannot process everything efficiently in working memory, thus inhibiting its transfer to long-term memory.

According to cognitive load theory, our brains can only take in so much information. The cognitive capacity of our working memory is limited to about seven items or elements of information at any one time [4]. Working memory is content-limited, such that we process a relatively small number of individual items or chunks at the same time. It is also time-limited, meaning we can only hold information in working memory for a very short amount of time [5]. So, if learning something new requires too much mental capacity, learning becomes hampered. Ideally, we want to optimize the use of working memory and avoid cognitive overload. Kirschner adds that we should design instruction so that working memory can process the information (content) that constitutes the instruction [6].

One approach to minimizing distractors while being mindful of cognitive load is to chunk content and information into short segments or units that actively engage the learner.  Active learning refers to an instructional approach that engages learners through meaningful activities where the learner is asked to participate actively by discussing, analyzing, questioning, creating, and/or applying concepts or skills in real time. These active learning chunks of content or information can then be molded into microlearning units. Microlearning may also be referred to as micro-eLearning, chunks, micro-media or micro media, bites, bite-sized learning, snacks, bursts, short courses, or micro-content [7, 8]. The term microlearning has no one conclusive definition, but several authors offer various descriptions as to what microlearning is [9 –11].

Torgerson and Iannone describe microlearning as a “mash-up of many things” that provides just-in-time learning support, which can be consumed quickly. They continue by saying that microlearning is content that can stand alone or support other learning activities [11].

Dolasinski and Reynolds explain microlearning as an approach “that focuses on a single concept, utilizing multisensory and multimodality in a focused short amount of time” [12]. They add that microlearning can be any learning content that stands alone or supports other learning activities, such as classroom instruction, online learning modules, etc. [12].

Kapp and Defelice define microlearning as “…an instructional unit that provides a short engagement [the exact number of minutes is not universally agreed upon] in an activity intentionally designed to elicit a specific outcome from the participant” [9].

Taylor and Hung claim the increased use of social media has changed the way people seek information, thus influencing their preference for single discrete topics presented in a short duration that meets their immediate information-seeking or learning needs [10]. Similarly, Mery posits that microlearning is more aligned with how today’s students consume new information, such as online TED Talks, YouTube videos, or text-based web pages [13]. The growing body of research around microlearning highlights its importance, particularly for Generation Z and future generations who have spent nearly their entire lives surrounded by computers, digital devices, and the world of social media [8]. There is a growing trend to have information available immediately and on demand. As such, students may expect faster and shorter learning experiences.

Microlearning can be offered in a variety of formats, such as video, podcasts, infographics, e-learning modules, or as structured text. While the length of time recommended for a microlearning unit is debatable and highly nuanced, the unit should be as long as it takes to address a single learning objective while engaging the student [11]. Others suggest that the content be “small nuggets of information, elements of informal learning, simple issues, and learning just-in-time on-the-job” [7].

Creating a Microlearning Unit
When it comes to creating microlearning units or segments, the evolving generative artificial intelligence (AI) tools (e.g., ChatGPT, Gemini, Claude) provide an efficient vehicle for microlearning development. By providing the AI tool with specific topics and prompts, content, and learning objectives, it can create concise, targeted information and generate quizzes, flashcards, and interactive scenarios in a matter of seconds. Using generative AI for creating microlearning units significantly reduces the time required to develop content. However, before using generative AI to create a microlearning unit, it’s important to consider the necessary elements of a microlearning unit.

To begin, Kapp and Defelice suggest writing one sentence as a concise overview about the content (targeted information), then elaborating and providing an example if needed [9]. There should be one clearly articulated learning objective for each microlearning unit, and each unit should address a specific concept or skill to ensure focus and clarity [7, 14]. Defelice also suggests the unit engage the learner in discerning, reflecting, and applying the content to real life [14]. Note, if the content is complex and cannot be meaningfully chunked out without losing coherence, it may not be suitable for microlearning.

The following example illustrates creating a microlearning unit for an online, graduate-level course using the AI tool, ChatGPT 4.o. ChatGPT is a natural language processing chatbot powered by generative artificial intelligence. Generative AI refers to algorithms that are capable of producing new content across various formats, including audio, computer code, images, text, and videos.

This example is taken from a course titled “Managing Learning in Organizations,” offered asynchronously for one quarter (10 weeks). The purpose of this microlearning unit was to quickly show students how to format citations and reference lists correctly, a requirement for all academic papers. The problem was that while students understood the need to cite sources, they consistently formatted citations and references incorrectly in their papers, often using a mashup of different formats, or barely formatting the reference list at all. Students were required to format their papers using The Chicago Manual of Style, 18^th edition, author-date format. This AI-generated microlearning unit was created to actively engage students in the process of correctly formatting in-text citations, as well as reference list entries for books, journal articles, and websites. It is recommended to start with one learning objective for the micro unit and let ChatGPT derive the content [14]. The objective in this example was “by the end of this microlearning unit, you will be able to accurately cite scholarly sources using the Chicago Manual of Style, author-date format.” The specific prompt given to ChatGPT 4.o was “Create a 5–10-minute microlearning unit that actively engages college students in learning how to correctly cite scholarly sources using The Chicago Manual of Style, author-date format.”  

Below is the microlearning unit ChatGPT 4.o created. At the end of the content or brief lesson, ChatGPT provides instructions and practice ideas for the instructor.

Learning Objective:
By the end of this microlearning unit, students will be able to accurately cite a scholarly journal article, book, and website using the Chicago Manual of Style, Author-Date format.

Microlearning Unit: How to Cite Scholarly Sources Using Chicago Manual of Style (Author-Date Format)

Introduction (1 minute)

Objective Overview: "In this microlearning unit, you'll learn how to correctly cite scholarly sources using the Chicago Manual of Style's Author-Date format. Let's explore the key elements of citations for journal articles, books, and websites."

Why it Matters: "Correctly citing sources is essential for academic integrity, allowing others to locate the resources you used and giving credit to original authors."

Part 1: The Basics of Author-Date Citations (1.5 minutes)

Explanation: A brief overview of what the Chicago Author-Date format looks like:

In-text citation: (Author Last Name Year, Page Number)

Reference list citation: Last Name, First Name. Year. Title of the Work, Publisher or Journal Name. URL or DOI if available.

Examples:

Journal Article: (Smith 2020, 150)

Reference list: Smith, John. 2020. "Title of Article." Journal Name 34 (4): 100-120. https://doi.org/xxxx.

Book: (Johnson 2018, 35)

Reference list: Johnson, Mary. 2018. Title of Book. City: Publisher.

Website: (Adams 2021)

Reference list: Adams, Sarah. 2024. "Title of Webpage." Website Name, April 15. https://www.website.com.

Part 2: Interactive Practice (3 minutes)

Task 1: Identify Citation Errors (2 minutes)

Instructions: Present three incorrectly formatted citations (one for a book, one for a journal article, and one for a website).

Activity: "Identify the error in the following citations. What needs to be changed to meet the Author-Date format?"

Example:

Incorrect: Smith, John. 2020. Title of article. Journal Name 34, no. 4: 100-120. DOI: xxxx.

Expected Answer: Add quotation marks to the title, format journal issue and volume properly, place DOI after period.

Task 2: Fill-in-the-Blank Citation (1 minute)

Instructions: Provide students with incomplete citations and ask them to fill in the missing elements.

Example: "Johnson, ____. 2019. Title of Book. ____, Publisher."

Feedback: Immediately show the correct answer, reinforcing correct citation elements.

Part 3: Quick Self-Assessment Quiz (3 minutes)

Question 1: What does an in-text citation look like in the Author-Date format for a direct quote from page 15 of a book by Jane Doe published in 2023?

Answer: (Doe 2023, 15)

Question 2: How would you cite an online journal article by Robert Black from 2019, titled "Climate Change and Urban Planning" in Journal of Urban Studies, found at www.journal.com?

Answer: Black, Robert. 2019. "Climate Change and Urban Planning." Journal of Urban Studies. www.journal.com.

Question 3: What key information is required for citing a website?

Answer: Author, year, title, website name, publication date, and URL

Delivery Tips:

Interactive tools: Use an interactive learning platform like Google Forms or a quiz tool to ensure real-time feedback and engagement.

Modality options: This can be delivered as a video, interactive slide deck, or through a learning management system (LMS) where students click through each section and complete tasks.

Embedded Resources: Include a downloadable Chicago Author-Date quick-reference guide for future use.

This AI-provided information and quiz can be quickly adapted for use in any online course, as suggested by ChatGPT 4.o. Once the correct answers are in place, students can refer to this unit on an as-needed basis.

As with any content generated from an AI tool, it is imperative to review the content for accuracy, tone, and word choice that reflects the instructor. Additionally, generative AI tools are known to hallucinate. “Hallucination" refers to the tendency of AI models to produce information that seems plausible but is factually incorrect or entirely fabricated. This can occur when the model generates content that doesn't exist or misrepresents the source material. To avoid hallucinating responses, be as detailed and specific as possible with the prompt, and double-check the response against a more reliable source, e.g., from databases such as ScienceDirect, Web of Science, or PubMed. Verifying information through multiple sources can help detect and correct any inaccuracies produced by the AI model. AI responses can sound very convincing, but they may be incorrect and in need of fact-checking. In the microlearning unit produced above, upon close review, I noticed several errors. In particular, the city or place of publication was included in the example; however, it is no longer required as of the 18th edition of The Chicago Manual of Style (2024).

Taking advantage of generative AI tools can make creating a microlearning unit quick and relatively easy. Upon reflection, the most challenging aspects were choosing appropriate content for the unit, deciding where it should reside in the online course, and where to provide links to the unit in other parts of the course, giving students access to the examples just in time, when needed.

Microlearning aims to address the learner's immediate needs by providing targeted information at the moment when the learner is ready to learn or needs the content. Microlearning units are designed to be completed in a short span of time and focus on a single skill, objective, or knowledge area, thereby reducing cognitive load. Creating engaging microlearning units can be expedited with generative AI tools that allow the instructor to focus more on other teaching and training endeavors. Generative AI tools can produce brief explanations, definitions, summaries, and worked examples relevant to the content. They also provide an effective and efficient way to create introductory content, summaries, quizzes, or knowledge checks. These all represent instructional practices that are characteristic of microlearning units. By focusing on specific, small chunks of content, microlearning helps improve knowledge retention, minimize cognitive load, and facilitate the application of learned concepts.

References

About the Author

Dr. Kim Hosler is a long-time learning and development leader with more than 30 years’ experience in learning leadership roles both in the corporate sector and in higher education.  She is passionate about developing effective and engaging learning and professional development initiatives that align with an organization’s business needs, vision, and mission. Hosler retired from the United States Air Force Academy, where she led course and curriculum development as their Director of Instructional Design. She currently enjoys teaching part-time at the University of Denver, in the College of Professional Studies. She earned her Master of Education (M.Ed.) from Kent State University and her doctorate in educational technology (Ph.D.) from the University of Northern Colorado.

This work is licensed under a Creative Commons Attribution International 4.0 License.