Enhancing the Comprehension of Science Text through Visual Analogies

Enhancing the Comprehension of Science Text through Visual Analogies Bryan J. Matlen (bmatlen@andrew.cmu.edu) Carnegie Mellon University, Department of Psychology 5000 Forbes Avenue, Pittsburgh, PA 15213 USA Stella Vosniadou (svosniad@phs.uoa.gr) National and Kapodistrian University of Athens, Department of Philosophy and History of Science University Town, Athens 161 21, Greece Benjamin D. Jee (bjee@holycross.edu) College of the Holy Cross, Department of Psychology 1 College Street, Worcester, MA 01610 USA Maria Ptouchkina (mptouchkina@u.northwestern.edu) Northwestern University, Department of Psychology 2029 Sheridan Road, Evanston, IL 60208 USA Abstract Instructional analogies are commonly used in science and mathematics text, yet students may have difficulty understanding analogies in the absence of adequate instructional support. In spatially rich domains like geoscience, visual depictions of both the base and target concepts of text analogies (i.e. visual analogies) may provide crucial support for students. To test whether visual analogies would be beneficial for learning, 72 fourth- and fifth-grade students were provided a short analogy-enhanced instructional text on plate tectonics that included either pictures of both the base and target concepts (Visual Analogy condition) or the pictures of the target concept only (Target Picture condition). Results indicated that children in the Visual Analogy condition outperformed children in the Target Picture condition on both near and far transfer measures. These results are consistent with recent research suggesting that factors that promote comparison – such as side-by-side presentation of examples – facilitate learning from text. Keywords: Analogy. Instruction. Science Cognitive Development. Geoscience Education. Education. Introduction Analogies pervade thought, thus, they are often used as scaffolds for student learning. Many studies have documented the effectiveness of analogies in math and science learning (Braasch & Goldman, 2010; Clement, 1993; Glynn & Takahashi, 1998; Iding, 1997; McDaniel & Donnely, 1996; Thompson & Opfer, 2010; Vosniadou & Schommer, 1988) and a review on the topic concluded that 12 of 15 studies showed positive effects for analogies in science education (Dagher, 1995). However, while there is general consensus that analogies support learning, substantially fewer studies have addressed how to optimize learning from analogies. Without adequate instructional support – such as guidance during the mapping process (Richland, Holyoak, & Stigler, 2004) or indications of “where the analogy breaks down” (Glynn, 1991) – students may fail to benefit from analogical comparisons (Richland, Zur, & Holyoak, 2007). The present study addresses what design characteristics lead to optimal learning outcomes from analogies. In particular, we assess how visual representations may be used to enhance analogical learning in elementary science education. Analogical comparison involves aligning two or more representations on the basis of their common relational structure (Gentner, 1983, 2010). When one of the analogs is better understood than the other – as is often the case in analogies used for instruction – information from the familiar case (by convention, termed the base) can be projected to the unfamiliar case (by convention, termed the target). These analogical inferences provide a powerful way to acquire new knowledge from prior experience. Research on analogy has revealed that factors that promote analogical comparison also facilitate relational learning (Brown, Kane, & Long, 1989; Christie & Gentner, 2010; Gentner & Namy 1999; Gick & Holyoak, 1983; Kotovsky & Gentner, 1986). For example, Camtrambone & Holyoak (1989) found that when college undergrads were prompted to compare two semantically dissimilar problems that shared a common solution, students were more likely to transfer the solution to a distant analog than students who received the same base examples without prompts to compare them. Recent research has demonstrated that comparison is effective for promoting learning in topics as diverse as mathematics (Rittle-Johnson & Star, 2009), biology (Gadgil, Chi, & Nokes, submitted), architecture (Gentner, Levine, Dhillon, & Poltermann, 2009), and business negotiation (Gentner, Loewenstein, & Thompson, 2001), and that even relatively mild manipulations that promote comparison – such as side-by-side presentation of examples – can facilitate relational learning (Christie & Gentner, 2010; Gentner, Loewenstein, & Hung, 2007; Loewenstein & Gentner, 2002; Oakes & Ribar, 2005). Although comparison in general has been found to promote learning across a diverse range of topics, the quality of the comparison can be an important factor in