The importance of causal antecedent emotional state inferences to narrative reading comprehension

Although research in narrative reading comprehension has supported the importance of inference processes to story comprehension, inference processes involving emotion have been studied only as elaborative, non-essential inferences. The purpose of the present project was to explore the contribution of a particular kind of emotional inference to narrative reading comprehension: causal antecedent emotional state inferences. The first, descriptive study tested for a positive association between appropriate causal antecedent emotional state inferences and narrative reading comprehension. The second, experimental study tested the hypothesis that appropriate causal antecedent emotional state inferences lead to narrative reading comprehension. Multiple-choice tests for both appropriate causal antecedent emotional state inferences and general comprehension level were constructed and used as dependent measures in both studies. In the experimental study, a "self-explanation" intervention was employed to promote appropriate causal antecedent emotional state inferences.;Analyses revealed a strong positive correlation between the causal antecedent emotional state inference questions and the reading comprehension questions; disattentuated correlations suggest that the two constructs may be identical as operationally defined here. A lack of significance in mean reading comprehension score between the experimental (self-explanation) and control (prediction) conditions in the second study, even using ITED reading comprehension scores as a covariate to decrease the error term, lends no support for the hypothesis that appropriate CAESI's lead to improved general comprehension. A lack of significance in mean number of causal antecedent emotional state inference score between the experimental and control conditions suggest a weak effect (if any) of self-explanation on the generation of appropriate causal antecedent emotional state inferences (as operationally defined here). It is hypothesized that improved operational definitions and measurement of both constructs---appropriate causal antecedent emotional state inference generation and reading comprehension level---along with a more specific comprehension intervention may reveal a significant relationship between the two constructs.