Fourth- and Fifth-Grade Argumentative Writing: An Analysis of Argumentative Reasoning and Connectives

• In Event: The Role of Texts in Dialogic Teaching

Abstract

Purpose
Our study examines cognitive and linguistic features in 4th and 5th grade students’ argumentative essays. First, we explore students’ uses of argumentative reasoning and uses of the linguistic feature of connectives (e.g., additionally, although) in their essays. Next, we test whether there is a relation between connectives and argument sophistication in the essays.

Theoretical Framework
This study draws from frameworks of functional linguistics, studies informing the development of argumentative reasoning, and a theory of general understanding of argument (Argument Schema Theory) in applied education settings.

Methods
The student texts were written on contestable topics as unit-culminating writing activities. We coded essays at the T-unit (Hunt, 1995) level according to Kuhn and Crowell (2011) argumentation scheme. We coded connectives according to Halliday and Hasan (1976) seminal categories. We report descriptive statistics for the argumentative and linguistic features used in the essays. We use regression techniques (with controls and adjusted standard errors for essays nested within students) to answer the question as to whether connectives are related to argument sophistication in the essays.

Data sources
The data were collected in 2013-2014 from 4th and 5th grades in two K-8 urban schools in the Northeast US as a part of a larger impact evaluation of Word Generation, a dialogic academic vocabulary program. Thirty-three 4th grade students and thirty-three 5th grade students wrote an essay on three separate topics per grade level (N=198 essays).

Results
Across the essay corpus, 86% of the essays contained at least one non-argument; 74% contained at least one own side only argument; 53% contained at least one dual perspective argument; and 27% contained at least one integrative perspective argument. Fifth grade essays contained significantly more own side only arguments (t = 5.29; p < .001) and more integrative perspective arguments (t = 3.91; p < .001) than fourth grade essays. Across the corpus, 57% of the essays contained at least one additive connective; 24% contained at least one adversative connective; 47% contained at least one causal connective; and 29% contained at least one temporal connective. Fifth grade essays contained significantly more additive connectives (t = 3.17; p < .01) and more adversative connectives (t = 3.35; p < .001) than fourth grade essays, but significantly fewer causal connectives (t = -3.60; p < .001) and fewer temporal connectives (t = -4.55; p < .001) than fourth grade essays. Regression analysis revealed a significant relation between adversative connectives (ß = .470; p < 001) and argument sophistication, as measured by the highest argument used in the essay.

Scholarly significance
Our findings suggest that brief writing exercises following classroom discussion and debate on weekly contestable topics may serve as a productive method for students to practice the micro-genre of argument/counter-argument within the macro-genre of persuasive writing. Our findings also suggest that these 4th and 5th grade students used specific word classes to express more complex arguments. Implications for practice, assessment, and future research will be discussed.

Authors

• Karen Taylor, University of California

• Joshua Fahey Lawrence, University of Oslo

• Carol M. Connor, University of California - Irvine

• Catherine E. Snow, Harvard University