Does Monitoring Event Changes Improve Comprehension?

Abstract

During narrative comprehension, reading times increase for changes in time, space, characters, goals, and causation. This study examined the extent to which instructional manipulations modify dimension monitoring during reading and whether this affects comprehension. Sixty-seven participants read three narratives (pretest). Half of the participants (the experimental group) were then instructed to simultaneously monitor changes in time, space, characters, goals, and causation. All participants then read three more narratives (post-test). After reading all texts, participants retold each story and answered comprehension questions. At post-test, the reading times for participants in the experimental group increased for changes in time, space, goals, and causation. Participants in the experimental group remembered more story information containing dimension changes compared with a control group (no instructional manipulation). However, participants in experimental and control groups did not differ in terms of overall memory or performance on comprehension questions. The results indicate that monitoring multiple dimensions simultaneously increases memory for those dimension changes but does not affect overall comprehension.

Acknowledgments

Special thanks to Melinda Mueller and Victoria Griffitt for their assistance in conducting this study.

Funding

This research was supported in part by a grant from Wichita State University.

Notes

¹ Individual difference measures were also collected and analyzed in a fourth step of the model. Curriculum-based measures of oral reading fluency (Deno, 1985) and a cloze maze task (using non-normed texts; based on the work of Deno [1985] and Espin & Foegen [1996]) were added to the level-2 models to examine cross-level interactions with the dimensions.

At pretest, participants with higher fluency and maze scores read the texts most quickly (fluency: γ₀₂ = − 8.97, p <  .001; maze: γ₀₃ = − 44.07, p = .008). There were no cross-level interactions between any of the dimension variables with the fluency or maze scores (p > .10). This suggested the fluency and maze scores did not interact with dimension monitoring.

At post-test, participants with higher fluency and maze scores again read the texts more quickly (fluency: γ₀₂ = − 8.30, p <  .001; maze: γ₀₃ = − 44.92, p = .002). There were no cross-level interactions between the dimension variables and fluency scores (p > .08). There were no cross-level interactions between maze scores and causal, goal, and character changes (p > .09). However, high maze scores decreased the effects of time (γ₁₁₃ = − 42.66, p = .007) and spatial changes (γ₁₂₃ = − 18.49, p <  .03) on reading times. This may indicate that individual differences in comprehension skill decreased monitoring. However, any conclusions drawn from the maze measure should be taken with caution because this measure has not been normed.

² To address individual differences, 2 (pre- vs. post-test, within subjects) by 2 (condition: experimental vs. control, between subjects) by 2 (high versus low scores on the maze or fluency tasks, as determined by a median split) ANOVAs were run, in which the dependent variable was the overall score on the comprehension questions. Participants who had higher maze and fluency scores performed better on the comprehension questions (F > 5.53, p <  .02). None of the interactions was statistically significant (F <  3.15, p > .08), indicating that monitoring did not interact with individual difference variables to affect comprehension.

³ To address individual differences in spoken recall, 2 (pre- vs. post-test, within subjects) by 2 (condition: experimental vs. control, between subjects) by 2 (high versus low scores on the maze or fluency tasks, as determined by a median split) ANOVAs were run. The dependent variables were the proportion recalled of clauses containing one, two, versus three simultaneous dimension shifts. In all models, the main effects of maze and fluency were not significant, nor were any of the interactions (F <  2.30, p > .11).