Publication Date


Document Type


First Advisor

Magliano, Joseph P.||Britt, Mary A.

Degree Name

Ph.D. (Doctor of Philosophy)

Legacy Department

Department of Psychology


Meaning (Psychology); Reading comprehension; Semantics--Psychological aspects


Readers must often construct an understanding of a complex process (e.g. climate change) from multiple documents. This is challenging because the documents are typically written for different purposes by different authors. In this context, readers must engage different reading strategies, because information that is important to the meaning of an individual text (i.e. text-based importance) may not be relevant to the reader's purpose (i.e. task-based relevance). Readers may have a default processing strategy based on attending to textual importance (i.e. importance-based processing), but when they have a goal or a task, they may switch to a processing strategy based on task relevance (i.e. relevance-based processing). Task-based relevance processing leads readers to allocate more attention to task-relevant content and less attention to task-irrelevant content, increasing memory for task-relevant information. However, task instructions can differ in the specificity of the cues they provide to help guide readers' identification and processing of relevant information. This may help explain why some types of tasks facilitate integration across multiple documents, while others do not. The more that relevance cues in task instructions guide readers to construct a mental representation dominated by interconnected, task-relevant content, the easier it should be for readers to reactivate and integrate information from prior documents. The current study investigated how the specificity of relevance cues in task instructions influence processing related to a causal explanation that could be extracted across documents. A low specificity task provided no relevance cues to guide processing. A moderate specificity task provided only semantic cues, and a high specificity task provided semantic cues and a structural schema cue (e.g., to help readers activate a causal explanation schema to guide processing). Two experiments investigated how these differences in tasks specificity influenced: (1) identification of text relevant to a causal explanation embedded across texts (Exp 1), (2) integration of text content during moment-to-moment processing (Exp 2), (3) content and organization of participant's recall (Exp 1 & 2), and (4) comprehension of the causal explanation (Exp 1 & 2). Greater task specificity was expected to lead to increases in outcomes related to the causal explanation across texts and decreases in processing of content that was only important to the individual texts. However, the possibility that attention to important text-based content would continue to influence processing was also considered. Experiment 1 showed that the addition of a causal schema cue in the high specificity task led to better memory for causal model content and increased integration across texts in participants' recalls. However, the semantic cue alone was sufficient to increase identification of text relevant to the causal model. There was no effect of task on causal model comprehension. In Experiment 2, the schema cue also led to better memory for causal model content. In contrast to Experiment 1, the semantic cue alone was sufficient to increase integration of causal model content across texts in participants' recalls. However, neither type of relevance cue affected moment-to-moment integration across texts. The schema cue, however, was more effective in increasing within-text causal model integration. Again, there was no effect of task on causal model comprehension. The results suggest that a semantic cue may be sufficient to help readers identify relevant text, but a schema cue was more effective for increasing memory and moment-to-moment within-text integration related to the causal model. However, the effectiveness of the schema cue on integration across texts in participants' recalls varied between experiments. Neither type of relevance cue influenced across-text integration during moment-to-moment processing, indicating a divergence between online and offline measures. Finally, while increased task specificity did shift processing towards the causal explanation content, it did not decrease processing related to important text-based content. Questions raised by these results, implications for models of task-oriented reading, and future directions are considered in more detail in the discussion section.


Advisors: Joseph P. Magliano; M. Anne Britt.||Committee members: Nicole LaDue; Michael Manderino; Keith Millis; Chris Parker.||Includes bibliographical references.


iv, 129 pages




Northern Illinois University

Rights Statement

In Copyright

Rights Statement 2

NIU theses are protected by copyright. They may be viewed from Huskie Commons for any purpose, but reproduction or distribution in any format is prohibited without the written permission of the authors.

Media Type