Navigating four billion videos: teacher search strategies and the YouTube algorithm

• Published in: Journal of educational media : the journal of the Educational Television Association Vol. 46; no. 1; pp. 47 - 59
• Main Authors: Fyfield, Matthew, Henderson, Michael, Phillips, Michael
• Format: Journal Article
• Language: English
• Published: Basingstoke Routledge 02.01.2021; Taylor & Francis Ltd
• Subjects: Algorithms; Catholic Schools; curriculum; Entanglement; Foreign Countries; Information Sources; Mathematics; Media Selection; new media; Online Searching; Pedagogical Content Knowledge; Search methods; Search strategies; Secondary School Teachers; Selection Criteria; teacher knowledge; Teachers; Teaching Experience; Teaching Methods; Video; Video Technology; Web 2.0 Technologies; YouTube; Australia
• ISSN: 1743-9884; 1743-9892
• DOI: 10.1080/17439884.2020.1781890

YouTube hosts a vast catalogue of instructional videos that are increasingly used in formal education contexts. Teachers regularly use YouTube to select videos for students, but the processes they use to select these resources have been understudied. This study explores how teachers search for videos, and the role of YouTube's complex algorithm in that process. It reports on a case study of nine Australian teachers working across two schools, at various stages of their careers, from a variety of subject domains. When searching for content on YouTube, most teachers use what can be described as a search and scroll method, meaning they enter a relatively simple search term then scroll through the resulting list. This strategy relies on the teacher to select appropriate videos displayed by YouTube's algorithm, representing an entanglement of human and software labour. The considerations of teachers around specificity to learning goals, length, engagement, and affect when choosing videos are discussed as important factors in final selection. While some teachers navigated through the search returns using well-developed content knowledge, others used rushed, or uncritical methods, increasing the influence of the algorithm in their practice.