[The crossover effect of the propositional text representation and the mental model:the role of individual differences in spatial imagery ability].

The crossover effect (Mani & Johnson-Laird, 1982) predicts that a mental model is more likely to be available after reading a determinate description of a spatial scene than after reading an indeterminate description whereas the reverse should apply to the propositional text representation. Although attempts to replicate this pattern were less than convincing, the crossover effect has often been stated as evidence justifying the differentiation of propositional representations and mental models. In an experiment with 44 adult, German-speaking participants, it was investigated whether the crossover effect is moderated by individual differences in spatial imagery ability. Participants read 6 determinate and 6 indeterminate descriptions of schematic, non-schematic, or schema-incompatible spatial scenes. Recognition data showed a crossover effect only in participants with lower ability. In the high-ability group the propositional text representation was less available than in the low-ability group, independent of whether or not a mental model could be constructed. The results explain why earlier replications might have failed. Consequences for current mental model theories of text comprehension are discussed.