Relational Scaffolding Supports Causally-Coherent Learning in Space Science

• In Event: 1-143 - Children’s understanding of counterintuitive scientific concepts: The role of coherent explanation-based instruction

Integrative Statement

Science learning involves understanding the structure of scientific theories (e.g., the heliocentric model of the solar system, the theory of evolution by natural selection, the particle theory of matter) and how these theories explain everyday phenomena. Achieving a coherent understanding of science is fundamentally a process of relational learning (Goldwater & Schalk, 2016): mapping the spatial, temporal, and causal relations between observations and their underlying explanation. The present study tested a method to support students’ causal-explanatory understanding of science.

We focus on a fundamental but counterintuitive topic in space science: the day-night cycle (e.g., Vosniadou & Brewer, 1994). Many children rely on Earth-based observations and use Sun motion to explain the phenomenon (e.g., Plummer, 2009). Further, although the cause of the day-night cycle is straightforward—Earth’s eastward axial rotation—it is not a trivial matter to explain how Earth’s eastward rotation causes an Earth-based observer to see the Sun rise in the east, occupy different locations in the sky at certain times of day, and set in the west. To help learners better understand the relationship between their Earth-based observations and a scientific model of the day-night cycle, we developed an intervention that involved relational scaffolding: guided visual comparisons between observed phenomena and corresponding modeled events. If learners struggle to understand science because of difficulties inferring connections between observations and a scientific model, then these guided comparisons should support their comprehension.

Third-grade children from public elementary schools (N = 99, Mage = 8.6) received 3 sessions of model-based instruction about the day-night cycle that included either (1) relational scaffolding—guided comparison between videos of Earth-based observations and modeled events or (2) sequential presentation—instruction involving the same video footage presented separately. All participants completed the same pre-, post-, and delayed posttest interviews that elicited causal explanations of the day-night cycle. Responses were coded using a reliable 27-item rubric (Cronbach’s α > .80) developed in prior research.

We conducted a multiple regression analysis to predict participants’ posttest understanding from pretest understanding, condition (relational scaffolding vs. sequential presentation), and a pretest x condition interaction term (controlling for age, gender, and spatial skill). The regression model accounted for 31% of the variance at posttest, F(11,87) = 3.52, SEE = 4.79, p < .001. The main finding was a significant pretest x condition interaction (β = -.41, p = .01, pr2 = .07). As shown in Figure 1 (left), relational scaffolding was especially effective for participants with relatively low (below-median) initial knowledge of the day-night cycle. At delayed posttest, the conditions diverged (see Figure 1, right) such that condition was a significant predictor (β = .30, p < .05, pr2 = .10), but pretest x condition was not (β = -.31, p = .10, pr2 = .04). Nonetheless, relational scaffolding had a greater impact on participants who initially expressed few scientifically-accurate ideas in their explanations. Thus, reducing the cognitive burden of integrating observations with explanations could help level the playing field for students who otherwise struggle to grasp causal-explanatory science knowledge.

Authors

• Florencia Anggoro, College of the Holy Cross
  Presenting Author

• Benjamin D Jee, Worcester State University
  Non-Presenting Author

• Amanda McCarthy, College of the Holy Cross
  Non-Presenting Author

• Victoria Jackson, College of the Holy Cross
  Non-Presenting Author