Analysis of Argumentative Writing at Two Grade Levels

Abstract The purpose of this study was threefold: (a) to investigate the effectiveness of four instructional strategies on student argumentative writing at the 10th- and 12th-grade levels; (b) to describe the differences among student writing at the two grade levels based on essays obtained at three different times; and (c) to determine how various qualities within the essays influenced the raters' decisions with holistic scoring. The results of the study indicate that there were no significant differences in the effectiveness of the four instructional strategies, which included instruction with model pieces of writing; with scales, questions, and criteria to guide writing; with models and scales/questions/criteria; and with free-writing activities. The results further indicate that the scoring of 10th and 12th graders' writing was not significantly different when using either a holistic scoring guide or Toulmin's (1958) model of argument. Both 10th and 12th graders, however, used significantly more argumentative traits (warrants and data) for one writing sample than for the other two samples. Results of a forward stepwise multiple-regression analysis indicated that approximately 30% to 45% of the variance of the holistic scores for the three writing samples was accounted for by the following variables—claims and data (for all three writing samples), opposition (for two writing samples), and warrant (for one writing sample). Overall, relatively few students included opposition or response to the opposition in their essays at either grade level. Their use of data and warrants was probably influenced by their knowledge about the topic. Findings of this study suggest that 10th and 12th graders' use of different components of an argument is relatively stable and differences may depend upon the amount of prior knowledge that the students have about the topic.