The construction of visual–spatial situation models in children’s reading and their relation to reading comprehension

•Spatial situation model construction during reading was tested in children.•Relations of spatial inference to reading decoding and comprehension were assessed.•Spatial information relevant to the protagonist’s perspective was most accessible.•Reading comprehension was uniquely predicted by inferred spatial location information. Readers construct mental models of situations described by text to comprehend what they read, updating these situation models based on explicitly described and inferred information about causal, temporal, and spatial relations. Fluent adult readers update their situation models while reading narrative text based in part on spatial location information that is consistent with the perspective of the protagonist. The current study investigated whether children update spatial situation models in a similar way, whether there are age-related changes in children’s formation of spatial situation models during reading, and whether measures of the ability to construct and update spatial situation models are predictive of reading comprehension. Typically developing children from 9 to 16years of age (N=81) were familiarized with a physical model of a marketplace. Then the model was covered, and children read stories that described the movement of a protagonist through the marketplace and were administered items requiring memory for both explicitly stated and inferred information about the character’s movements. Accuracy of responses and response times were evaluated. Results indicated that (a) location and object information during reading appeared to be activated and updated not simply from explicit text-based information but from a mental model of the real-world situation described by the text; (b) this pattern showed no age-related differences; and (c) the ability to update the situation model of the text based on inferred information, but not explicitly stated information, was uniquely predictive of reading comprehension after accounting for word decoding.