Schema utilization in readers of varying abilities

An attempt was made to replicate and extend Mandler and Johnson's (1977) study in which simple stories were parsed according to a model story grammar system and the recall protocols obtained from subjects were analyzed for omissions and errors. Mandler and Johnson reported that the pattern of omissions and errors obtained during recall could be predicted on the basis of story grammar considerations. The results of the present study indicated that processing variables such as the mode in which the story is presented to subjects as well as subjects' level of reading comprehension ability can affect the pattern of omissions and errors obtained from subjects during recall. These results are not well accounted for by models that stress the importance of story structure as a prime determinant of recall. A more specific scoring system than that presented by Mandler and Johnson was developed for this study, and this system made it possible to examine the nature of the observed processing differences more closely. Subjects who listened to a taped version of the story before recall had a difficult time recalling the story. They tended to add material that acted as "filler" at places where their memory for the actual story was weak. There was a slight tendency to add or replace propositions to make the story more meaningful at times during recall. Subjects who read a printed version of the story did not have as much difficulty recalling the story and tended to condense individual story propositions into summary statements during recall. When propositions were added or changed, the better readers tended to make changes that added to the cohesion or meaningfulness of the story more often than did readers of lesser ability. Better readers were hypothesized to have more sensitivity to both grammatical structure and content of story material and to be superior at implementing inferential processes to organize recall than readers of less ability.