Skip to main content

Andrew Taylor ACF Abstract FY11

"Balanced evidence processing: Evidence based and affect-based subjects process scientific texts differently"

Conference Name: Conference of the Society for Text and Discourse

Processing and memory for scientific texts were examined as a function of belief towards a topic and the basis for that belief (evidence vs. affect). Sentence reading times were recorded while subjects read texts that either supported or refuted specific topics (evolution and television violence). Processing was assessed by regressing reading times on certain sentence characteristics, including the extent to which each sentence supports or refutes the text position. Belief basis, but not the beliefs themselves were a good predictor of reading times. Evidence-based subjects slowed down their processing for sentences that opposed the main text position (e.g. pro-evolution sentences in a text that argues against evolution). No such trend was seen in affect-based subjects. Results suggest that evidence-based subjects will, more than affect-based subjects, alter their level of processing throughout a text to create a balanced situation model.