YouTube as a learning resource: a review and case study in AI

YouTube was created in 2005 as a video-sharing website and since has evolved to become the biggest such platform. Among all the different content, it also includes some pedagogical and didactic videos on different disciplines.

This post is adapted from a project I wrote during my PhD for the Academic Teacher Training course and comes from my personal experience as a learner during my university years. I made a habit of watching YouTube videos before the exams, as I found it a perfect way of summarizing the knowledge acquired during the lectures and as a reminder for some concepts. The fact than anyone can make a video, means that usually for a specific subject there are several sources, and thus ways of presenting the same information. Of course, a video allows to use all kinds of animations and other resources to explain abstract concepts in an engaging way.

This post will start by analyzing the current state of YouTube as a teaching and learning resource and end with a case study of Artificial Intelligence. Hopefully, this can be used to critically analyze how is YouTube used as a dissemination tool of Artificial Intelligence and how can it be used in one’s own teaching.

Some insights from the literature

Below is a non-extensive table showing the number of results in Google Scholar for different word combinations. As expected, the number decreases as more words are added.

Keyword	N. of results
YouTube	6,440,000
YouTube + Learning	2,100,000
YouTube + Teaching	1,360,000
YouTube + Learning + Teaching	695,000
YouTube + Learning + Teaching + Artificial Intelligence	60,200

Early studies highlight the potential of YouTube as a teaching resource

When using only “YouTube” as the keyword without specifying years, some of the most relevant studies shown in Google Scholar included pedagogical and academic papers. As early as 2011, there already are very in-depth reviews of YouTube across many disciplines ¹, as well as analyses of YouTube as an educational tool ². They already start highlighting some of the obvious potentials such as using the video as a model for later discussions or helping teachers with their planning, but they also identify copyright and credibility as challenges.

This search also yielded some examples of a trend that continues thought this study and it is the fact that the medical field has always been analyzing in this context. ^{3 4}.

Newer studies focus on misinformation

When filtering for papers from 2018 on, many papers were about the radicalizing potential of YouTube ⁵ and medical misinformation on topics such as COVID-19 ⁶, prostate cancer ⁷ or brain cancer treatment ⁸.

There is a shift, then, from a tool used to aid teachers, to a tool that students should use independently, so their personal responsibility becomes a bigger factor.

YouTube is used for solving doubts

Probably the most complete and most cited survey regarding YouTube as a teaching and learning resource was made by Moghavvemi et al in 2018 ⁹. That paper presents that indeed students use YouTube for solving their academic doubts. This findings are shared by other researchers ^{10 11}.

Another reference study by Orús et al ¹² presents how the generation of videos by the learners themselves enhance cross-curricular aptitudes and help students to engage in active learning

Medical fields dominate the case studies

Searching for “Youtube + learning”, many results included case studies, with medical and biological themes dominating ^{13 14 15 16}. Other themes like performing arts ¹⁷ and language learning ^{18 19}, mainly English as a second language ^{20 21}, also appear, the latter being more prevalent when filtering by recent articles.

There were not many research papers that included all the keywords including “Artificial Intelligence”. There was only one paper directly assessing the quality of YouTube videos about Artificial Intelligence, but it was even more specific, about dental radiology ²², concluding that even though there were not many videos on the subject, they were of acceptable quality.

Other articles broadly analyzed Information Technology ²³ or Science, Technology, Engineering and Maths (STEM) ²⁴. They both agreed that using these resources enhances motivations and understanding in students, as well as participation, in line with the findings in other fields.

Case study: AI videos on YouTube

I chose three videos by searching on YouTube “Artificial Intelligence”, “Deep Learning” and “Machine Learning”. The videos that were an entire lesson or documentary of more than 20 minutes were excluded, as this work focuses on YouTube-specific features. The criteria for choosing the videos will be (1) number of visualizations, as a measure of how popular the video is with the general audience and (2) my own familiarity with the channel, as a measure of how popular the video is with people on the field.

Title	Views	Uploaded date	Channel name	Subscribers
What Is Artificial Intelligence? - Crash Course AI #1	536,897	9 Aug 2019	CrashCourse	13.6M
How Machines Learn	7,398,838	18 Dec 2017	CGP Grey	5.52M
But what is a neural network? - Chapter 1, Deep Learning	11,733,544	5 Oct 2014	3Blue1Brown	4.65M

The CrashCourse video is the introduction of a 20 video course on Artificial Intelligence. The course seems very complete covering main aspects of the subjects. Undoubtedly, a good resource for students to check and get a different view on concepts they might have been taught during lessons. That said, for the sake of making it more fun and dynamic, the video does show a robot, which is very stereotypical when speaking about Artificial Intelligence and can lead someone completely alien to the topic to have an early association that not many teachers would agree with.

The same happens with the CGP Grey video, that uses robots to explain some of the points. This video is absolutely intended for viewers that are interested in Artificial Intelligence but not necessarily students. However, they present the topic professionally and there are no evident wrong facts stated in the video.

Finally, the 3Blue1Brown video does exactly the opposite and is clearly intended for the students that have not yet understood how neural networks work, a very fundamental concept for Artificial Intelligence. They constantly use animations and examples that make a very clear explanation for someone at any level in their journey.

---

I would like to give a special mention to my colleagues Erik Hallström, Daniel Gedon and Fredrik Gustafsson that prepared a video with accompanying blog posts in Educaora for 7th to 9th grade students explaning the main concepts of machine learnings and giving a specific example for heart attack discovery from patient’s electrocardiograms.

Discussion and conclusions

YouTube as a teaching and learning tool has been researched from as early as 2011. YouTube was a different platform then and the way internet was consumed has varied in these 10 years. However, some researchers already saw the potential of using YouTube to help students in the classroom.

Interestingly, the focus shifted from studying how the teachers could use YouTube for students to how students were using YouTube to help them in their learning process. Thus, it became more and more important that the quality of the videos was appropriate for self-consumption, as now teachers would not have the control of what pedagogical content the students were consuming.

The most studied field for using YouTube as a learning resource is Medicine and Healthcare. And for good reason, as it is a delicate matter which any person, student or not, may consume and also many videos are created as compared to other fields. Thus, even if there is good quality medical content in YouTube, one has to be careful for the stated facts.

Other disciplines are much less student comparatively, but also the relative amount of videos is lower. This also makes the few videos that are created of overall better quality.

From the case study, this seems to be the case also for Artificial Intelligence. The analyzed videos were intended for somewhat different audiences, but all of them conserved a good quality level and are surely used by confused students to clarify concepts.

Teaching application

Artificial Intelligence videos on YouTube are generally trustworthy and could be recommended to students for further consultation. These videos could be included at the end of the lecture for the students to watch at home or even shown during classes -or special seminars- where the students could further ask questions about the contents of the videos.

There is also potential for asking the students to make videos themselves explaining the concepts they feel most comfortable with and sharing them both with the class and with the general public, as an assignment.

YouTube as a teaching tool is more about empowering the students to correctly use it, than for teachers to show the videos on the classroom.

References

1. C. Snelson, “Youtube across the disciplines: A review of the literature,” MERLOT Journal of Online learning and teaching, 2011 ↩

2. T. Jones and K. Cuthrell, “Youtube: Educational potentials and pitfalls,” Computers in the Schools, vol. 28, no. 1, pp. 75–85, 2011 ↩

3. K. C. Madathil, A. J. Rivera-Rodriguez, J. S. Greenstein, and A. K. Gramopadhye, “Healthcare information on youtube: A systematic review,” Health informatics journal, vol. 21, no. 3, pp. 173–194, 2015. ↩

4. B. Drozd, E. Couvillon, A. Suarez, et al., “Medical youtube videos and methods of evaluation: Literature review,” JMIR medical education, vol. 4, no. 1, e8527, 2018 ↩

5. K. Munger and J. Phillips, “Right-wing youtube: A supply and demand perspective,” The International Journal of Press/Politics, vol. 27, no. 1, pp. 186–219, 2022 ↩

6. H. O.-Y. Li, A. Bailey, D. Huynh, and J. Chan, “Youtube as a source of information on covid-19: A pandemic of misinformation?” BMJ global health, vol. 5, no. 5, e002604, 2020 ↩

7. S. Loeb, S. Sengupta, M. Butaney, et al., “Dissemination of misinformative and biased information about prostate cancer on youtube,” European urology, vol. 75, no. 4, pp. 564–567, 2019 ↩

8. K. ReFaey, S. Tripathi, J. W. Yoon, et al., “The reliability of youtube videos in patients education for glioblastoma treatment,” Journal of Clinical Neuroscience, vol. 55, pp. 1–4, 2018 ↩

9. S. Moghavvemi, A. Sulaiman, N. I. Jaafar, and N. Kasem, “Social media as a complementary learning tool for teaching and learning: The case of youtube,” The International Journal of Management Education, vol. 16, no. 1, pp. 37–42, 2018 ↩

10. E. T. Maziriri, P. Gapa, and T. Chuchu, “Student perceptions towards the use of youtube as an educational tool for learning and tutorials.,” International Journal of Instruction, vol. 13, no. 2, pp. 119–138, 2020 ↩

11. Y. Chtouki, H. Harroud, M. Khalidi, and S. Bennani, “The impact of youtube videos on the student’s learning,” in 2012 international conference on information technology based higher education and training (ITHET), IEEE, 2012, pp. 1–4 ↩

12. C. Or ́us, M. J. Barl ́es, D. Belanche, L. Casal ́o, E. Fraj, and R. Gurrea, “The effects of learner-generated videos for youtube on learning outcomes and satisfaction,” Computers & Education, vol. 95, pp. 254–269, 2016 ↩

13. A. Clifton and C. Mann, “Can youtube enhance student nurse learning?” Nurse education today, vol. 31, no. 4, pp. 311–313, 2011 ↩

14. A. Raikos and P. Waidyasekara, “How useful is youtube in learning heart anatomy?” Anatomical sciences education, vol. 7, no. 1, pp. 12–18, 2014 ↩

15. S. C. Burke and S. L. Snyder, “Youtube: An innovative learning resource for college health education courses.,” International Electronic Journal of Health Education, vol. 11, pp. 39–46, 2008 ↩

16. T. Akgun, C. Y. Karabay, G. Kocabay, et al., “Learning electrocardiogram on youtube: How useful is it?” Journal of electrocardiology, vol. 47, no. 1, pp. 113–117, 2014 ↩

17. D. DeWitt, N. Alias, S. Siraj, M. Y. Yaakub, J. Ayob, and R. Ishak, “The potential of youtube for teaching and learning in the performing arts,” Procedia-Social and Behavioral Sciences, vol. 103, pp. 1118–1126, 2013 ↩

18. J. Brook, “The affordances of youtube for language learning and teaching,” Hawaii Pacific University TESOL Working Paper Series, vol. 9, no. 1, p. 2, 2011 ↩

19. M. Alhamami et al., “Observation of youtube language learning videos,” Teaching English with Technology, vol. 13, no. 3, pp. 3–17, 2013 ↩

20. H.-c. Wang and C. W.-y. Chen, “Learning english from youtubers: English l2 learners’ self-regulated language learning on youtube,” Innovation in Language Learning and Teaching, vol. 14, no. 4, pp. 333–346, 2020 ↩

21. R. Kabooha and T. Elyas, “The effects of youtube in multimedia instruction for vocabulary learning: Perceptions of efl students and teachers.,” English Language Teaching, vol. 11, no. 2, pp. 72–81, 2018 ↩

22. K. Cesur Aydin and H. G. Gunec, “Quality of information on youtube about artificial intelligence in dental radiology,” Journal of Dental Education, vol. 84, no. 10, pp. 1166–1172, 2020 ↩

23. L. Abdillah, “Enriching information technology course materials by using youtube,” in The 5th International Conference on Artificial Intelligence, Computer Science and Information Technology (AICSIT2017), Bayview Beach Resort, Batu Ferringhi, Pulau Pinang, Malaysia, 2017 ↩

24. W. O. Otchie, M. Pedaste, E. Bardone, and I.-A. Chounta, “Can youtube videos facilitate teaching and learning of stem subjects in high schools?” Bulletin of the Technical Committee on Learning Technology (ISSN: 2306-0212), vol. 20, no. 1, pp. 3–8, 2020 ↩