The Short- and Long-Term Effect of Explicit Grammar Instruction on Fifth Graders' Writing
Feng, Shoudong, Powers, Kathy, Reading Improvement
Grammar instruction has long been a troubling issue for many language arts teachers. This collaborative research between an elementary classroom teacher and university faculty, based on the assumption that grammar is most effectively taught in reading and writing, looks into the short- and long-term effect of error-based grammar instruction on the writing of a group of fifth grade students. Results indicate that mini-lessons that target errors identified in previous students' writing produce positive short- and long-term effect on students' rewriting. An instructional model based on the findings of this study is also proposed.
The need for grammar instruction and the method of teaching it in elementary language arts curriculum have long been controversial issues (Cramer, 2004; Tompkins, 2002). Hillocks and Smith (2003) conclude with evidence from their meta-analysis that focusing on instruction is much more effective in improving student writing than on grammar and mechanics. Their findings provide support for some of the earlier classic research on this topic (Elley, Barham, Lain, and Wyllie, 1976; Harris, 1962). On the other hand, Weaver, McNally and Moerman (2001) believe that to teach or not to teach grammar is not the question; it is a matter of what and how to teach it. They strongly oppose the isolated teaching of grammar rules and concepts, which is actually what their opponents investigated in their research and found to be ineffective. Putting the debate aside, a close look at samples of elementary students' writing suggests that students in reality frequently make grammar mistakes in writing, and many of the mistakes are consistent and should be addressed. At present, more and more researchers and educators seem to support grammar teaching, but only in the context of reading and writing (Cox, 1999; Cramer, 2004; Patterson, 2001 ; Weaver, McNally, & Moerman, 2001; Tompkins, 2002). Furthermore, many researchers specifically recommend teaching it in the revising, proofreading, or editing stage (Cox, 1999; Sharon, 1997; Patterson, 2001; Weaver, 1998).
What seems to be missing in the debate, however, is a model that delineates the specific procedures in the error-based approach and the systematic evidence that this approach is in fact effective, both in short- and long-term, on the accuracy of student writing. The present research, which is a collaborative effort of a classroom teacher and college faculty, is intended to fill the gap by measuring the result of this kind of instruction. Therefore the questions we ask in this research are: 1) Does error-based grammar instruction have positive short-term effect on student writing? 2) Does error-based grammar instruction have positive long-term effect on student writing?
The approach employed in this study is comprised of analysis of grammar errors in student writing, minilessons that target those errors, and reanalysis of errors in the follow-up writing. Positive effect is defined as reduced numbers of grammar errors in student writing. For this study, a decrease of 20% would be considered positively effective. Conclusions will be drawn on the effect of this method of instruction.
Participants: The participants of this study are a group of fifth graders in a public elementary school in a southern state. Their numbers in different stages of the study vary slightly depending on whether the students were willing to participate at that particular time. At the beginning of the study, twenty two students participated, but the number went down to sixteen halfway through and to nineteen at the final stage of the study.
Data collection and analysis: The data of this study consist of student writing samples collected at three different points in the school year. The first batch of samples was collected at the beginning of the school year, on the topic of "My Friends", assigned by the classroom teacher. …