Teaching Text Structure = Better Essay Content

Our take? Although the research cited below shows that strong writers are able to identify and duplicate text structures in essays they’ve read, it also shows that those who can identify and duplicate these structures in their own writings tend to add more detail and information than those who don’t.

Lesson learned? Teach students text structures to advance reading comprehension and written expression.

Title: Structure and content in eighth‐graders’ summary essays
Authors: Golden, J; Haslett, B; Gauntt, H
Source: Discourse Processes 11, 139-162 (1988)
Link: https://psycnet.apa.org/record/1989-24041-001

Abstract: This study was meant to evaluate eighth-graders’ summaries of a scientific article. Revealed were several qualitative differences among students’ summaries. However, the primary purpose of the study was to develop a model for analyzing expository text that would take into account both the superstructure (text organization) and the macrostructure (semantic content— ideas and meanings expressed in sentences that pertain to the logic) of expository writing. A second purpose was to examine the qualitative differences between summary essays written by students, and a third purpose was to assess the relationship between the students’ summary essays and the input essay they were summarizing.

Overall, the higher-skilled summaries were more detailed, identified the dominant predicate more completely, and used a variety of rhetorical structures compared to the lower-skilled summaries. With regard to organization, the higher-level essays reflected the text organization of the original text (the input essay). The study revealed that students were attentive to the text organization and semantic content (that is, content relating to meaning in language or logic) of the input essay, and drew upon these structures to construct their summaries.

Participants/Studies Included: The study was made up of 37 eighth-grade students from an East coast middle school. The input essay that was used was “Treating Yourself for Minor Aches and Pains,” from Science World magazine.

Research Design: Qualitative mixed-method experimental study

Methods: Students were asked to read an article and then to summarize it in their own words (students were permitted to refer to the article while writing their summaries).

After gathering the summaries, they were analyzed using Meyer’s system, which was divided into two categories: one for structural rhetorical predicates and one for rhetorical predicates that reflected content. Meyer’s system is used to study the relationship between text structure and the reader’s comprehension and recall of text; it maintains that “the structural hierarchy of a text is identified on the basis of logical relations” and that rhetorical predicates (such as collection, causation, response, comparison, and description) specify the relationships among propositions and order them into hierarchical relationships. “A top-level, superordinate structure organizes the overall text and subsumes all the content and relationships in the text.” Additional categories (developed by van Dijk), which reflected superstructures of orientation and contextual information, were added to Meyer’s system.

Meyer’s System: The top section contains the different types of text structures. The bottom section contains the different meanings or ideas expressed in sentences contained in the body of essays.

Three raters coded the superstructure (text organization) and the macrostructure (semantic content) of the input essay, and they then independently coded each of the students’ written summaries of the article. Three scores were assigned to each summary: one score for the superstructure, one for the macrostructure, and one for the global score (the combined superstructure and macrostructure). Samples were then assigned to three levels based on the global score, ranging from well-formed summaries to less well-formed summaries; higher level summaries (Level III essays) scored from 20 points and above, middle level essays (Level II essays) scored from 10 to 19 points, and lower level essays (Level I) scored from 1 to 9 points.


Results: In the low-level summaries, only 3 out of 11 writers identified any key idea, only a little over half used orientation (scene-setting), and none identified the top-level structure, meaning that the problem/solution structure of the input essay was not used in summary organization. The average writers, however, consistently identified key ideas. Most used orientation, and all writers successfully identified the problem/solution structure. The average summaries were also far more detailed than the low-level summaries. The high-level summaries contained significantly more key ideas than the other two groups, and all summaries recognized the problem/solution structure of the input text and used it to organize their own summaries. They were also far more detailed compared to the low and even the average-level summaries. Most of the high-quality summaries contained orientation or contextual material.

Implications: The investigators have provided a method of essay analysis that reveals much about students’ writing. Via this analysis, we find answers to questions that can guide instruction. Such questions are:

  1.  Can students identify the text structure of the essays they are to summarize?

  2. Can students duplicate these structures in their own writing?

  3. What semantic ideas are students able to express and which do they not use? 

The authors express their concerns regarding the impact on students of poorly written texts whose structures are unclear.

Leave a Comment