Methods and Measures for Using Writing to Transform Knowledge in Science Classes

Abstract

This dissertation documents three connected studies addressing critical issues in writing-to-learn research: a) how to measure students’ feelings about writing, b) how to assess scientific writing, and c) how to integrate writing-to-learn into current secondary science curriculum. Considered in concert, this work seeks to provide measures and methods for using writing as a tool to transform knowledge in secondary science classes.

Students’ emotions about writing impact achievement, and therefore tools capturing motivation, attitude, and self-efficacy are needed to determine the extent that writing achievement is a result of skill development, affective issues, or a combination of both. Specifically, my first study describes the validation of a newly developed measure of self-efficacy towards writing for middle grades students called the Student Writing Affect Survey (SWAS). Findings indicate the SWAS yields reliable and valid scores to measure middle grades students’ self-efficacy towards writing.

The purpose of the second study was to create and validate a rubric, known as the Rubric for Scientific Writing (RSW), which can be used to support writing instruction in science classes and evaluate scientific writing. This rubric assesses both students’ general writing skills and their ability to write appropriately within the scientific genre. My findings demonstrate that the RSW produces valid and reliable scores for two factors of students’ scientific writing – scientific argumentation and English rhetoric. The RSW has the potential to aid both science teachers who may lack training in the teaching and assessment of writing as well as researchers who need a stable measure of students’ scientific writing.

Finally, the third study uses these tools to measure the effectiveness of a writing-to-learn intervention in middle and high school science classes. Prior literature posits that writing-to-learn strategies are less effective for younger students; however, few studies have implemented similar strategies across grade levels. Therefore, this study combines established best-practices to create a writing-to-learn intervention that can be implemented into existing science classes at various grade levels. While high school students did slightly outperform their middle-grade peers, further cluster analysis demonstrated that students who created visuals and used scientific vocabulary during the intervention made the most growth, regardless of grade level.