Publication Date


Document Type


First Advisor

Britt, Mary A.||Millis, Keith K.

Degree Name

Ph.D. (Doctor of Philosophy)

Legacy Department

Department of Psychology


Science--Study and teaching; Learning; Psychology of; Comprehension--Psychological aspects


Scientific passages are important because they provide information about how the world functions, and therefore, are ubiquitous in school contexts and necessary for STEM education. Because of their importance, it is crucial to identify and study factors that impede their comprehension so that educators know how to help readers understand them more deeply. Indeed, past research on the comprehension of science texts has shown that they are typically difficult to comprehend (Graesser, 1981; Millis, Graesser, & Hamberlandt, 1993). Given the difficulty of scientific passages, it would be informative to understand how readers create a coherent representation of the text while reading by connecting ideas within the text. One way to assess readers' coherence building is by examining whether they notice when information in the text is inconsistent. Readers often do not consciously report problems in scientific passages (explicit detection) (Glenberg, Wilkinson, & Epstein, 1982; Noordman, Vonk, & Kempff, 1992; Otero & Kintsch, 1992). Therefore, researchers have tried to examine coherence building during reading. They found that, at least under optimal circumstances, readers do spontaneously generate inferences while reading (Singer & Gagnon, 1999; Wiley & Myers, 2003). These findings provide indirect evidence that readers are detecting breaks in causal coherence, as bridging inferences can be generated as a means to repair breaks in coherence during reading (Otero & Kintsch, 1992). However, no research has provided direct evidence that readers detect breaks in causal coherence while reading science texts. The aim of the current dissertation therefore is to examine (1) whether readers spontaneously detect coherence breaks within scientific explanations as measured by reading times (implicit detection), (2) whether implicit detection depends on reading skill or task instructions, and (3) whether measures of implicit detection and conscious awareness of a problem (explicit detection) converge. Participants read short passages presenting a scientific explanation of a phenomena (e.g., how honey is formed) that included an inconsistency between the events of the causal chain. Both implicit (reading times of target statement when inconsistent vs consistent) and explicit detection (rating of how inconsistent the target statement was when inconsistent vs consistent) were examined. The results suggest that regardless of reading skill or task, participants took longer to read the target sentences in the inconsistent condition than the consistent condition. However, probing questions about awareness of the inconsistencies indicated that regardless of task or reading skill, participants were not aware of the inconsistencies. These results suggest that readers detected the break in coherence created by the conflicting words within the sentence (e.g., cooled-heated), but are not detecting the break in causal coherence between the events of the explanation. In addition, detecting a coherence break between conflicting words while reading does not lead to explicit awareness that an inconsistency had been encountered.


Advisors: Mary A. Britt; Keith K. Millis.||Committee members: Amanda M. Durik; Joseph P. Magliano; Michael L. Manderino; Patrick A. Roberts; Katja Wiemer; Jennifer Wiley.||Includes bibliographical references.


ix, 122 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