This study investigated sketch maps and written summaries as measures of large-scale spatial representations as wellas learning aids to improve navigation proficiency. One hundred and fifty-six participants explored a virtual environment(VE) comprising independent and connecting routes. Participants were then asked to sketch or write a summary describingthe layout of the VE. A free exploration phase followed in which they could learn more. The testing phase comprised twoobjective measures of navigation proficiency: a pointing task and a model-building task. Sketches provided significantlymore target and route details about the VE than written summaries, although the quality of both correlated with objectivenavigation measures. Thus, both are good measures of spatial representations, despite prior doubts about them. However,neither sketching nor written summaries positively influenced subsequent exploration or spatial learning. Symbolic rep-resentations may not be effective tools for improving navigation skills. Another possibility is that they may be but onlywith further represent-explore-feedback cycles.

Many people experience illusions of understanding for explanations of scientific phenomena (Rozenbleit & Keil,2002) and readers tend to be poor at gauging how well they have understood what they have read in expository science texts(Dunlosky & Lipko, 2007; Maki, 1998; Thiede, Griffin, Wiley, & Redford, 2009). The present line of research includesstudies demonstrating that metacomprehension accuracy may be especially poor when students are presented with texts thatinclude features such as diagrams, graphs, animations, and analogical examples. Although these adjuncts are meant to improvecomprehension, they can often lead to illusions of understanding. An important theme of this research is articulating the kinds ofinstruction and skills that students may need before they can learn effectively from expository science texts including graphicsor analogies.