Intellectual merit.

In our first three years, we have developed and pilot tested five new instructional techniques aimed at teaching high-school science students how to effectively use the diagrams that appear in their own biology textbooks. We gathered eye-tracking data on students’ real-time behavior while learning from diagrams, and developed two new measures of diagrammatic reasoning.

 

Research from cognitive psychology, educational psychology, and other disciplines has amply documented the difficulties that people have making sense of diagrams (Graesser, Lu, Olde, Cooper-Pye, & Whitten, 2005; Hannus & Hyona, 1999; Kozhevnikov, Motes, & Hegarty, 2007 Kriz & Hegarty, 2007; Mayer, 2005; Paas, Renkl, & Sweller, 2003). In our own research, we have found that students simply skip large portions of diagrams in biology text (Cromley, Luciw, & Snyder, 2008). We have also found that inference and other high-level processes that are important for learning from text are even more important for learning from diagrams (Cromley, Snyder, & Luciw, 2009). Students typically show low gains in conceptual understanding from existing diagrams in textbooks, but few interventions have ever been developed and tested to improve classroom instruction about how to use diagrams (e.g., Seufert, 2003; Bodemer & Faust, 2006). Science textbooks include a wide range of images, including line diagrams, naturalistic drawings, flow charts, chemical diagrams, and hybrid diagrams (e.g., a photograph with a schematic diagram) (Pozzer & Roth, 2003). Our research suggests that line diagrams are both the most prevalent and the most complex images in biology textbooks (Cromley et al., 2009). In a high school textbook we found that line diagrams had a mean of 4.2 features per image versus photographs (M = 1.3 features per image) and other image types (e.g., flow charts, M = 2.3 features). Line diagrams are therefore a logical place to design interventions to help students learn to use images in science text.

 

We have capitalized on commonalities among several theories of diagrammatic reasoning to develop five interventions: 1) conventions of diagrams (COD; i.e., teaching students how to read captions, color keys, etc.), 2) COD + coordinating text and diagrams (CTD), 3) COD + CTD + self-explanation, and 4) COD + CTD + student-completed diagrams—verbal completion and 5) visual completion. Together with pre- and post-intervention measures of background knowledge and diagrammatic reasoning, we have also collected data on spatial ability and visuospatial working memory in order to assess the effect of these individual difference characteristics on learning how to better understand scientific diagrams. In conjunction with these experimental studies, we have and will continue to collect eye tracking data on a subset of participants while they learn from diagrams. Eye tracking data collected pre-intervention—i.e., time spent looking at diagrams vs. text, transitions between text and diagrams, and which relevant features of diagrams are inspected (i.e., Regions of Interest)—establishes a baseline for gaze patterns of more-proficient students compared to less-proficient students. Eye tracking data collected pre-and post- intervention shows how students’ gaze patterns change with intervention.