Cognitive Overload or Motivational Anchor? Evaluating AI-Integrated 5E Science Instruction in Grade 7 and 8 Classrooms via a Quasi-Experimental Crossover Design

Abstract

Concerns regarding decreased attention spans associated with modern digital media consumption challenge traditional direct teaching[cite: 8]. This study evaluates a multimedia-integrated teaching method on Grade 7 and 8 students at Miza Educational Institute in Hulhumalé, Maldives[cite: 8]. Using a quasi-experimental crossover design, 115 experimental observations were gathered from 91 unique students, with participants crossing over to experience both the experimental and control conditions[cite: 8]. The intervention included an AI-generated video hook, direct instruction, a summary video, and an immediate retrieval quiz[cite: 8]. Analyzed via linear mixed-effects modeling (LMM), results revealed a profound developmental interaction effect driven by intrinsic cognitive load[cite: 8]. For Grade 7 (low intrinsic load), multimedia integration acted as an effective motivational anchor, yielding a statistically significant improvement in immediate retrieval scores[cite: 8]. Conversely, for Grade 8 (high intrinsic load), the data suggests that the intersection of high element interactivity and the transient information effect likely induced cognitive overload, acting as the primary explanatory hypothesis for the performance drop[cite: 8]. In this context, the multimedia appeared to act as seductive details, resulting in significantly higher immediate retrieval scores for the traditional direct instruction cohort[cite: 8]. AI-generated multimedia must be rigorously calibrated to learners' cognitive maturity and the syllabus's intrinsic difficulty[cite: 8].

Keywords

• Educational Technology
• Science Education
• Middle School Education
• Adolescent Learning
• 5E Instructional Model

How to cite