Analytics-supported Collaborative Learning: Impact on Students of Different Cognitive Abilities and from Different Socioeconomic Backgrounds

Dr Chen, Gaowei   (Principal Investigator (PI))

Professor Chan Carol Kwai Kuen   (Co-Investigator)

18

2017-04-01

2018-09-30

77570

Analytics-supported Collaborative Learning: Impact on Students of Different Cognitive Abilities and from Different Socioeconomic Backgrounds

Analytics-supported collaborative Learning, Classroom videos, Learning analytics, MOOCs and Flipped Classroom, Video observational learning

Others - Education

201611159070

Seed Fund for PI Research – Basic Research

2016

Completed

Purposes and objectives The proposed project aims at investigating the impact of a new learning approach that leverages technology in learning analytics to support students’ collaborative review of videos. In this approach, students observe videos of lectures that they have previously participated in and perform problem-solving tasks. Students’ video observation will be supported by a video-observation platform, which offers the state-of-the-art visualization and analytics of classroom talk while automatically synchronizing the video and transcripts. By transforming traditional means of tutorial work or homework largely by textbook review to the review of classroom lecture videos, this new learning approach is hypothesized to significantly enhance students’ learning outcomes in science, especially for those who do not afford time to engage in deep thinking and reasoning during classroom time and for those who cannot afford extra after-class private tutoring. The project has the following objectives:  to examine the effects of the new approach on middle school students’ learning and development in science;  to understand the interactions between the new learning approach and students’ individual differences in cognitive abilities;  to evaluate the impact of the new approach on students from different socioeconomic backgrounds. Key issues and problems being addressed The well-known prevalence of private supplementary tutoring (also called shadow education) in Asia is causing mounting financial pressure on parents (Bray, 2009, 2013) and positions students from low socioeconomic backgrounds at disadvantage. The development of electronic resources such as Khan Academy and digital libraries is promising for moderating this uneven social problem by providing self-study resources. Instructional videos are gaining increasing popularity in learning environments such as Massive Open Online Courses (MOOCs) and flipped classrooms (Chen & Wu, 2015; Hew & Cheung, 2014; Fernandez, Simo, & Sallan, 2009; Guy & Marquis, 2016; McGarr, 2009; O’Flaherty & Phillips, 2015). A review of the literature has demonstrated that students collaboratively observing dialogic videos were most effective for improving learning outcomes in comparison to other video-based learning conditions such as observing monologue videos and observing videos alone (Chi, Kang, & Yaghmourian, 2016; Chi, Roy, & Hausmann, 2008; Craig, Chi, & VanLehn, 2009). However, researchers also found that it is challenging for young learners to benefit from the videos as they may easily lose focus when the video content is overwhelming (Muldner, Lam, & Chi, 2014). Findings from a few more recent studies are inspiring for the potential solutions to the problems regarding short attention span during students’ observation of videos. It was found that the relevance of videos to the in-class lectures was especially conducive to the successful implementation of instructional videos for learning, due to the connections across learning situations and space (Fößl, Ebner, Schön, & Holzinger, 2016; Muñoz-Cristóbal et al., 2014). Students were more likely to review the concepts and issues that they had learned in lectures as their prior knowledge from the lectures encouraged their effective use and acceptance of videos, and their assimilation of new materials in the video (Lonn & Teasley, 2009; Jiménez-Castillo, Sánchez- Fernández, & Marín-Carrillo, 2016). The benefits from revisiting the learned materials in computer-mediated learning demonstrated above were consistent with the Knowledge-Learning-Instruction framework, which regards learning not only as conceptual understanding, but also for fluency building and refinement through practices (Koedinger, Corbett, & Perfetti, 2011). Therefore, to help middle school students learn effectively from existing school resources, we develop a new learning approach called Analytics-supported Collaborative Observation of Classroom Videos (ACOCV), which leverages advances in learning analytics, and particularly through embedding Classroom Discourse Analyzer (CDA; Chen, Clarke, & Resnick, 2014, 2015) into a video-observation platform to present classroom videos as reusable resources to students. In this way, it replaces the prevailing linear formats of video presentations in video-based learning environments with multiple visualizations (e.g., temporal and graphic representations of classroom talk, transcripts, and the synchronized video segments), to reduce learners’ cognitive load for non-learning-related activities during video observational learning, and afford them greater autonomy in selecting video segments of interest for viewing and reviewing. Therefore, the new approach promises to solve the problem of short attention span during young learners’ video observation, and boost their learning efficiency and effectiveness. Research questions and hypotheses The proposed project will investigate the new analytics-supported collaborative learning approach and test its effectiveness on students’ learning in science, and collect first-hand evidence on its impact on students from different socioeconomic backgrounds and of different cognitive abilities. The primary three research questions of the project are: 1) Does the new analytics-supported learning approach help Hong Kong middle school students learn science effectively and why? 2) Does the new approach impact on students of different cognitive abilities differently? 3) Does the new approach impact on students from different socioeconomic backgrounds differently? Regarding the first research question, it is hypothesized that the new learning approach will significantly improve students’ learning outcomes in science in comparison to students who observe videos without the analytics support. The second and the third research questions will examine how students' individual differences in cognitive abilities (e.g., prior knowledge in science, and their working memory span) and their family socioeconomic backgrounds are related to their learning outcomes. It is hypothesized that learners with lower prior knowledge, and shorter working memory span will benefit more from the analytics-supported learning approach as there may be ceiling effects for students of higher cognitive ability (i.e., they do not need the review and revisit opportunities). It is likewise that students from lower socioeconomic backgrounds will benefit more from the new learning approach as they have less alternative resources (e.g., private tutoring after school) available.