EFFECTS OF TEXT AND FEEDBACK VARIANTS ON SPATIAL INFERENCING

Two versions of a fictitious recreational park were developed. One des cribed the location and order of 25 park landmarks as might be encount ered by a driver (route version). The other (survey version) presented the same 25 landmarks using spatial or geographic referents. Both ver sions were 25 sentences in length. Subjects read either the route or s urvey Version first, one sentence at a time. After reading the text, s ubjects were given I min. access to a map of the park. The route subje cts were given a label map, which showed the location of park landmark s by name. The survey readers were given an icon map, which designated the same landmarks using pictures. After viewing the map, both groups were given a set of 20 inferential questions regarding the landmarks. Subjects then read the same text version again; how ever, half the su bjects for each text version viewed the same map, while the other half received the alternate map form. All groups then answered the same se t of inferential questions. A control group received no feedback on ei ther reading. Analysis showed that, compared to the control group, the subjects with feedback showed greater gains in scores across trials. Survey text readers showed greater gains than route Version readers. F or both versions, readers given access to the alternate map showed gre ater gains than those who viewed the same map twice. Moreover, the sur vey readers benefited more from the map change than the route readers. The data support prior research which suggests that the survey text b ase appears more flexible in terms of making greater use of available feedback, even if different from the original.