Abstract HTML Views: 514 PDF Downloads: 142 Total Views/Downloads: 859
Abstract HTML Views: 364 PDF Downloads: 104 Total Views/Downloads: 607
We investigated the effect of spatial memory capacity and content maps on readers' memory for hypertext structure. Simple hierarchical hypertexts were built on four topics. For each topic there was a 6-node and a 9-node version. Each version came with or without a content map. Young adult participants were asked to read each hypertext with the purpose of learning the contents and structure. Then, they had to recall the layout of nodes and links. Memory for links varied as a function of spatial memory and the presence or absence of a map. When no map was available, high spatial memory participants drew more accurate maps than low memory participants. When a map was available the two groups had better results and did not differ from each other. The benefit of the map was larger for 9-node than for 6-node items. The results indicate that mentally representing hypertext structure relies on spatial working memory. Global content representations act as scaffolds for low spatial memory users.