Searching for Explanations in Science Trade Books: What can we learn from Coh-Metrix?

Given the importance of explanation in science texts for children, this research explored Coh-Metrix as a tool for rapidly determining the presence of explanations in science trade books. Variables from Coh-Metrix specifically addressing cause (causal content, causal cohesion, and positive causal connectives) were correlated with an existing, high-inference causal variable from previous research. Moderate correlations were found for causal cohesion and positive causal connectives (both r(18) = 0.46, p < 0.05); no correlation was found with causal content. Somewhat stronger correlations were found with two, co-referential Coh-Metrix variables, argument overlap and stem overlap. These findings suggest that Coh-Metrix may be used for an initial screening to determine if a science text has the necessary conditions to be explanatory.