Knowledge representation as a function of test expectancy
The present study examined the effects of test expectancy (verbatim or inference) and instructions concerning the reality of story information on knowledge representation. College-level students read a story and then participated in a sentence verification task which included either verbatim or inference statements from the story. A context effect was found for the inference, but not for the verbatim statements. That is, subjects who received verbatim statements tended to incorporate the story information, udiile subjects who were presented with inference statements were more likely to compartmentalize the story information. Furthermore, subjects responded faster to the inference statements than to the verbatim statements. It was expected that subjects would display a context effect uhen given fictional story instructions and no context effect when presented with real story instructions. This result was not obtained, however. A structural model does not accurately predict the outcome of the present study as to the effects of statement type. Such a model would predict compartmentalization in subjects presented with verbatim statements and incorporation in subjects presented with inference statements.