Figurative Language in Computer Education: Evidence from YouTube Instructional Videos

Abstract

Educators are utilizing both traditional and novel electronic methods in teaching, including web-based modalities. According to existing literature, Web 2.0 in general and YouTube videos specifically are increasingly utilized in education. Web 2.0 applications, such as the popular online video network YouTube may enhance the students' learning and retention while providing connections with peers and faculty. The use of online video platforms, including Youtube in education, has been widely discussed in various research studies, including YouTube's role in the students' learning support, motivation, and engagement. Additionally, researchers have studied the use of YouTube videos in students' self-regulated learning. However, few studies focused on the instructor's stylistics in educational videos. Stylistics is the study and interpretation of texts or spoken language regarding their linguistic and tonal style. In this work, we studied instructor stylistics with a focus on the use of figurative language. In engineering courses, instructors tend to use metaphors, similes, and other forms of figurative language to connect new concepts with students' existing knowledge base or communicate complex concepts with various non-technical audiences. Unfortunately for researchers interested in understanding instructional processes, most of these lectures or discussions are not recorded nor have footprints on the internet, making it hard to analyze and understand how instructors use this type of language in teaching. We utilized YouTube as a rich data source containing millions of educational videos that can be beneficial for qualitative and quantitative studies that allow the opportunity to discover interesting patterns and themes. Specifically, We report on the results of a study in which we searched for, collected, and analyzed the use of figurative language in educational videos on YouTube teaching topics related to computer education. The videos are related to different topics in Computer Science and are targeting audiences of varying age groups. This approach is a multi-method one that involves both qualitative and quantitative data analyses. We present the data collection and analysis pipeline used in our study as a potential tool for other researchers to implement in their work. The generated dataset from the pipeline consists of video transcripts, metadata, and other statistical and linguistic video attributes. Overall, this research work provides two contributions: (i) a view of how instructors utilize various forms of figurative languages across different topics and age groups in computer education and (ii) a fruitful approach to educational data collection. The results from the data analysis provide an opportunity for pedagogical and linguistic guidance to engineering education researchers and faculty members on the use and integration of accessible figurative language within the classroom. The generated dataset can also support linguists and computer scientists working on language, speech, and audio processing software systems and services, particularly conducting research within the fields of human-computer interaction, human-human interaction, video content understanding, and interactive dialog systems.