Knowledge acquisition from newspaper stories∗

Abstract

Two experiments tested subjects' acquisition of the information found in newspaper stories. Subjects studied either an actual newspaper story or an altered form of the story constructed by deleting the irrelevant and redundant information and restructuring the remaining information into either a narrative, topical, or outline form. Subjects who studied the intact newspaper story learned little of the irrelevant and redundant information compared to the other information in the story. Presentation of any of the condensed versions of the story led to better recall than presentation of the entire news story. However, none of the structural reorganizations of the text information led to consistently superior recall relative to a passage that deleted the irrelevant information but maintained the news story structure. These results suggest that people may select among several alternative structural schemata for the representation of narrative passages in memory.The influence of text organizations on the acquisition of information from texts has received extensive treatment by educational and cognitive psychologists. Several studies have investigated the efficacy of various information organizations for learning. In a series of related studies (Frase, text passages were constructed from matrices of name-attribute-value triples. That is, a set of concepts in a semantic category was described using a fixed set of attributes, and each concept in the category had different values of the attribute. In general, these studies demonstrated that coherent organizations of the information (i.e., either by names or by attributes) produced better learning than did a random organization and that the two organizations produced equivalent levels of learning and better learning than a random organization. While name organization was a preferred or dominant learning strategy, attribute organization could be adopted if the to-be-learned texts were so structured.

Constructing inferences in expository text comprehension

“…These elements, in turn, are organized into abstract structures, such as linear chains and hierarchies; and into rhetorical networks, such as the comparison of two or more elements (Black, 1985). Research concerning expository text comprehension has yielded evidence supporting structures of this sort, such as the linear chain (Seely & Long, 1994;Singer, 1982;Thomdyke, 1979). In general, however, if expository text can be characterized as schematic in form, then the relevant sché-mas must be particularly flexible and dynamic in character (Miller, 1985).…”

Initial mention as a signal to thematic content in technical passages

“…For example, Carpenter and Just (1977) postulate a "discourse pointer" whose initial state is determined by the initial portion of a passage. From theories of prose comprehension based on schema ideas (e.g., Thorndyke, 1977Thorndyke, , 1979, we would expect that the main subject should appear early in the passage to enable the reader to activate the relevant portions of his general knowledge or to perform the initial step of selecting the proper schema. Others point out that initially mentioned information may be required to provide a contextual framework for the passage content (e.g., Bransford & Johnson, 1972) or to define the major passage referents (Clark & Haviland, 1977).…”

Text processing and knowledge updating in memory for radio news∗

“…For instance, van Dijk (1980, p. 121) has proposed a preliminary account of a schema for newspaper articles consisting of three main parts, Introduction (which also functions as Summary), Specification, and possibly various Details that may be cut by the editor if there is lack of space. Thorndyke (1979) assumed a similar schema to be usual in news. He went on to compare the organization of stories found in newspapers with different ways of organizing the same contents, e.g., a chronological one (called "narrative organization")-He found effects on both reading time and recall, but organization interacted with the contents of the stories, thus suggesting that just one schema is not sufficient.…”

“…This field was initially explored by Bartlett (1932), but systematic theorizing and experimentation had to await the development, in linguistics and artificial intelligence, of tools for describing the structure of texts and events. Presently, several theories are available and a large number of empirical studies have been carried out, mostly concerning short, fictional stories in the tradition of the oral folktale (e.g., Kintsch, 1977;Mandler & Johnson, 1977;Rumelhart, 1975;Thornydyke, 1977); but also magazine articles (Meyer, 1975), research reports (Kintsch & van Dijk, 1978), lectures (Kintsch & Bates, 1971), and newspaper articles (Thorndyke, 1979) have been investigated.…”