No abstract

No abstract

Waern, Y. Structure in similarity matrices. A graphic approach. Scand. J. Psychol., 1972, 13, 5–16.—The present paper proposes a simple method for representing the structure inherent in the similarity matrix, i.e. a graphic representation. The traditional multidimensional scaling methods are compared with each other and with the graphic representation with regard to representation, interpretation and assumptions concerning the subject's performance. The graph gives a topological representation and the derivation of the graph from the similarity matrix may be made so as to reduce the demands on the subjects somewhat. As to interpretation, the graph is able to reveal dimensions, clusters, and hierarchies. Compared with some multidimensional scaling it gives very similar interpretations.

The application of the think-aloud method to the study of different text processing tasks is illustrated and discussed. Exploratory studies indicated that some stable general characteristics of think-aloud comment during reading could be found. These characteristics could be categorized reliably in different studies of the reading of various texts. The suggested categories were also sensitive enough to capture the effects of different instructions and texts. In later studies, think-aloud comments were used to illuminate conditional reasoning and Interpretation of words in context. The thinkaloud comments formed the basis for suggesting a descnptive model for the latter task.To conclude, it was found that the think-aloud method forced the consideration of a wide ränge of aspects in describing text processing. Whereas previous models have mainly focused on intratextual relationships and readers* prior knowledge, the new aspects include the use ofprior knowledge not only for comprehension but also for idea processing and criticism.The think-aloud method further Stresses the importance of considering evaluation and monitonng processes. In particular, readers have to decide how to continue their processing, once they have failed to comprehend. Finally t the reconstruction of meaning adds a creative aspect t o the pure text processing.

Two empirical studies are presented, investigating the effect of readers' world knowledge on comprehension of a text (assimilation effect) and the effect of the reading of the text on readers' knowledge of the world (accommodation effect). The definition 'knowledge of the world' is here restricted to a part, represented by the readers' beliefs. 46 junior college students and 1 1 university students served as subjects. The results regarding assimilation showed that the average subject in both studies interpreted the text to have a meaning more similar to his own beliefs than was actually the case. As to accommodation, the junior college students' own beliefs changed in the direction of the text, whereas the university students' own beliefs became even more adamant in opposition to the text. '" 11 IS Scand. J . Psychol. 18 132 Y . Waern