ENG 574: Prior knowledge, reading skill, and text cohesion in the comprenhension of science texts - Sarah Felicelli

Article Citation
Ozuru, Yasuhiro, Kyle Dempsey, and Danielle S. McNamara. 2009. “Prior Knowledge,
Reading Skill, and Text Cohesion in the Comprehension of Science Texts.” Learning and
Instruction 19: 228-242.

Summary of article
Ozuru, Dempsey and McNamara looked at the influence of reader characteristics (prior knowledge and reading skill) and text characteristics (text cohesion and text difficulty) on reading comprehension of a biology text. They selected an excerpt from a biology text and created a high-cohesion version and a low-cohesion version. Students from two different universities participated in the study.

All participants were tested to determine their level of reading skill and their level of prior knowledge of the subject prior to reading the text. Comprehension questions consisted of text-based questions (directly from information in a sentence), local-bridging questions (student had to link information found “within five clauses…(generally adjacent sentences)”), and far- or global-bridging questions (students had to link information that was “more than five clauses apart, and more than two sentences apart”) (232-233).

Researchers learned, for text-based questions, reading comprehension was higher when a high-cohesion text was read before the test. However, there was not a positive correlation between text cohesion and comprehension for bridging questions. Ozuru, Dempsey, and McNamara believe the amount of a reader’s prior knowledge combined with the difficult nature of science text led to this result. A reader with little previous knowledge of science focuses more on understanding the information at the text level and less on trying to make inferences between different topics in a document (239).

For readers with low reading skill, however, the researchers found that a high-cohesion text can actually hinder their ability to recall content, because the length of the document also typically increases. A longer text increases the difficulty level for students with low reading skill, which leads to decreased comprehension.

Analysis and Reflection
This is a recent study which was published two years ago. The data analysis used in this study seems appropriate for the mixed model design that was used (2X2X2X3 - text cohesion X level of knowledge X reading skill X types of comprehension questions). The researchers used ANOVAs to analyze the data. They also used regression analysis; however, I am not as familiar with that statistical analysis, so I am unable to determine if that was appropriate or not. The reliability (Cronbach's alpha) of the reading skill test and the topic-specific knowledge test questions were both high (0.90 and 0.73 respectively); however, the reliability for the biology knowledge questions used in the test was only 0.61, which seems rather low. It would be interesting to see if increasing the reliability of the biology knowledge questions would have any effect on the results of the study.

Again, this seems to be an appropriate article for my research project as it addresses the influence of text cohesion (which includes syntax and semantics) on the comprehension of science text.