Using analogies to improve elementary school students' inferential reasoning about scientific concepts

School Science and Mathematics, Mar 2001 by Yanowitz, Karen L

Various scientific concepts were taught to students in the third through sixth grades. Some children were taught the concepts using instructional analogies. Each analogy explicitly compared the science concept to a more familiar topic. Other children received expository texts not containing analogies. Students were asked to recall the texts and to answer inference questions about the science concepts. Fourth- and sixth-grade students read the texts on their own in Experiment 1. Students who read the analogical text showed higher levels of performance on inference questions than students who received the non-analogical texts. In Experiment 2, texts were read aloud to third- and fifth-grade students. The analogical texts were read once, and the nonanalogical texts were read twice to equate the number of times students were exposed to the general principles governing the domains. As in Experiment 1, students who received the analogical texts demonstrated better inferential reasoning than students who received the non-analogical texts.

Students often find learning science to be an especially challenging endeavor. Using analogies as explanatory devices can be a useful way to teach science (Glynn, 1997; Kolodner, 1997; Lawson, 1993; Thagard, 1992; Venville & Treagust, 1997), and teachers often use analogies when explaining scientific concepts (Dagher, 1995). An analogy is formed between two concepts when structural relations present in both concepts are placed into correspondence (Brown, 1989; Clement & Gentner, 1991; Gentner, 1989; Holyoak & Koh, 1987). Individual elements in the two domains might be semantically quite different, but the structural relations between those elements would be the same in both domains. For instance, a common instructional analogy compares the flow of water through a hose with the flow of current through a wire (Hewitt, 1987). Although water and electricity are different elements, the general principles governing the way electricity flows through pipes and the way water flows through hoses are similar. Therefore, students can use their knowledge of the structural principles governing water flow to help them understand electricity.

After similar structural relations are placed in correspondence and the analogy is formed, constructive use of the analogy may occur. Students can use their understanding of the principles described in a more familiar context (often referred to as the source domain) to comprehend a relatively unfamiliar (or target) domain. Several recent studies with college students have examined how using analogies to teach various science concepts influenced inferential reasoning about those concepts.

For instance, Halpern, Hansen, and Riefer (1990) examined college students' understanding of three different science concepts as a function of reading instructional analogies. Instructional analogies explicitly compare two domains by describing the similarities in their structure (Newby, Ertmer, & Stepich, 1995; Stepich & Newby, 1988). Some students read short paragraphs that contained analogies, such as comparing the movement of lymph through the body to the movement of water through the spaces in a sponge. Other students read nonanalogical expository texts about the science concepts. Participants were asked factual questions about information directly presented in the texts, as well as inferential questions requiring them to use their understanding of the domain to generate the answer. Participants who read the analogical texts showed a higher level of performance on both the factual and inferential questions compared to those who read the nonanalogical texts. Using similar designs, Iding (1993) and Donnelly and McDaniel (1993; McDaniel & Donnelly, 1996) foundthat college students who learned scientific concepts using analogical texts were also able to answer significantly more inferential questions compared to those who had received nonanalogical texts.

The research with college students clearly shows that using an instructional analogy to teach an unfamiliar science domain increased students' inferential reasoning about that domain. The present research was designed to examine whether or not elementary school students would also show such a benefit. Many studies have shown that elementary school students are capable of analogical reasoning. Little prior research, however, has specifically explored the effect of analogies on inferential reasoning in scientific domains with younger students.

The majority of the studies examining analogical reasoning in elementary school students have used a problem solving situation (e.g., Brown, 1989; Goswami, 1991). The typical problem solving paradigm presents a narrative source story, which explicitly states how the protagonist of the story solved a problem. The student can then transfer this given solution principle to solve the target problem. In contrast, the instructional analogies described previously never explicitly provided the answer to inferential reasoning questions. Presumably, answering inferential questions is a more difficult task, or atthe very least a different task, than using information directly provided in a source domain. Therefore, it is important to examine whether or not elementary school students would also benefit from receiving instructional analogies when engaging in inferential reasoning tasks.