Writing To Learn: A Mixed Methods Case Study of Reflection/Exit Writing in Fourth Grade

Development in writing, motivation to write, and student self-perceived writing ability all play pivotal roles what students are able to produce. The National Commission on Writing in America’s Schools and Colleges (2003) placed writing at the center of educational reform, calling upon the educational system of the nation to participate in a “writing revolution.” Data support this call to arms: the National Assessment of Educational Progress reported that only 23% of fourth graders wrote at the “proficient” or "advanced" levels; the majority of children—61% of fourth-grade students—wrote at the “basic” level; 16% of fourth graders produced “below basic” writing (U.S. Department of Education, 1999). To address these concerns about the state of writing in America, this study investigates a content-area writing intervention, Reflection/Exit writing, and its effect on three student writing outcomes: (1) student self-perceptions; (2) writing development as measured in quantity; and (3) writing development as measured in quality. Freiberg (1993) developed Reflection/Exit writing to help teachers establish a calm, productive end to class, bring closure to their lessons, and enable purposeful reflection on the learning for the day the learning of the day during the last five to six minutes of class, by asking students to reflect on what was learned that day. This mixed-methods case study included a sample of 56 fourth grade students, in both bilingual and traditional (ESL) classrooms, in a predominately Hispanic, low SES elementary school. Two intervention classrooms taught by fourth grade Math/Science teachers and two comparison classrooms taught by fourth grade Language Arts/Social Studies teachers were the units of analysis and multiple points of data were examined for each classroom. A concurrent, parallel mixed-methods design was employed, utilizing qualitative and quantitative methodologies, which were analyzed through three different strands of research. In Research Strand 1, samples were analyzed for compositional fluency, or length, by calculating the number of words and syllables to determine if students were able to produce a greater quantity of writing over time. Research Strand 2 was used to determine if the intervention affected the quality of student writing over the study period through the use of the state's holistic writing rubric (used from 2003-2011; The Texas Education Agency, n.d.) and through content analysis procedures. The holistic rubric considered writers’ focus and coherence, organization, and development of ideas. Content analysis procedures assessed writers’ cognitive development in writing, through the themes of: (1) planning; (2) knowledge telling; and (3) knowledge transforming (Flower and Hayes, 1981; Bereiter & Scardamalia, 1987). In Research Strand 3, students' writing self-perceptions were measured through the Writer Self-Perception Scale (Bottomley, Henk, & Melnick, 1997/1998). Results indicate that when implemented with fidelity, Reflection/Exit writing improved the quality of student writing, as measured through content analysis and scale scores on the state's holistic writing rubric. In the case of the high-fidelity intervention, improvements in writing quality from pre-post intervention, as measured on the holistic rubric, were significant (p = .002), with a large effect size (ɳ2 = + .54; see Cohen, 1998). Comparison group classrooms and the low-fidelity intervention classroom did not have significant gains in the quality of writing from pre-post intervention. Between groups (intervention vs. comparison), there was a significant difference between students’ change in writing quality, as measured on the holistic writing rubric (p = .005; ɳ2 = + .17). Students in the high-fidelity intervention group also demonstrated improved writing quality through content analysis measures, with higher levels of cognitive development in writing at post-intervention. Comparison classrooms and the low-fidelity intervention classroom made little growth in cognitive development in writing. Comparison classes observed statistically significant gains in the length of the writing samples from pre-post intervention, as did the high-fidelity intervention classroom. Between groups, however, there were no statistically significant differences in the change in writing length. There were also no statistically significant differences in students’ writing self-perceptions in either comparison or intervention classrooms. This study demonstrates that when Reflection/Exit writing is implemented with fidelity, students in the intervention classroom outperformed comparison group students in writing quality on the holistic rubric and in their levels of cognitive development in writing. When the intervention was implemented without consistency or fidelity, there were no notable changes in student writing quantity, quality, or self-perceptions. This study sets an important precedent—student growth in writing should be analyzed through multiple lenses and from various ways of knowing. Implications for this study include the expanded use of Reflection/Exit types of writing to improve the quality of student writing. Preparing for a post-secondary-ready environment builds at the early grades; writing skills are a necessary building block for future success (National Commission on Writing in America’s Schools and Colleges, 2003). Most students can write adequately, but few can write at a high degree of proficiency (Institute of Education Sciences, 2008). This study implies there is a need for content-area teachers (e.g. Math/Science teachers) to dedicate a few minutes each day writing about what students learn across the curriculum in order to improve writing quality. Future research should examine the use of Reflection/Exit writing with bilingual-only populations, as an intervention for LEP students, as well as its expanded use with different ages of learners.