ENG 574: Reversing the Reverse Cohesion Effect: Good Texts can be Better for Strategic, High-Knowledge Readers - Sarah Felicelli

Article Citation
O’Reilly, T. and Danielle S. McNamara. 2007. “Reversing the Reverse Cohesion Effect: Good Texts Can Be Better for Strategic, High-Knowledge Readers.” Discourse Processes 43(2): 121-152.

Summary of the Article
In a 2007 study, Tenaha O’ Reilly and Danielle S. McNamara evaluated the validity of the reverse cohesion effect. The reverse cohesion effect refers to the increased level of comprehension exhibited by high-knowledge readers when they read a low-cohesion text (122). O’Reilly and McNamara used a high- and a low-cohesion text—an excerpt from a high school biology text on cell mitosis (cell division)—to study the comprehension for students at Old Dominion University. For the high-cohesion texts, the excerpt was revised by:

(a) replacing pronouns with noun phrases,
(b) adding descriptive elaborations,
(c) adding sentence connectives,
(d) replacing or inserting words to increase conceptual overlap,
(e) adding topic headers,
(f) adding theme sentences, and
(g) moving or rearranging sentences. (8).

O’Reilly and McNamara discovered that only high-knowledge readers with less reading skill exhibited the reverse cohesion effect. Those readers remembered more after reading a low-cohesion text than after reading a high-cohesion text. Based on the results of their study and how the interaction between prior knowledge, text cohesion, and reading skill affect comprehension of material, they recommend that writers and editors compose documents that are highly cohesive while keeping in mind the amount of prior knowledge that a reader may have (24).

Analysis and Reflection
The sample size for this study was much larger than the other three I have posted so far. The researchers had 143 students who were participants in this study. Again, the data analysis was appropriate for the design. The researchers used a mixed design (2x2x2x2) and analyzed the data with ANOVAs and correlations to determine relationships between question type, text type, comprehension skill and science knowledge. The article is a continuation of research that was begun with the 1996 article by McNamara et al. This will also be an article that will be useful for my research project for ENG 574.