The Effect of Past Learning Experience on Writing Performance in an Upper‐Level Undergraduate Science Course

Abstract

Acknowledging the influence of students’ prior learning experiences, understanding the connections of this prior learning to current learning contexts, and designing learning experiences for transfer of content to new contexts is essential for student achievement of expertise in a discipline (Bransford, Brown & Cocking, 2000). Writing in the discipline is an important skill for both developing and demonstrating disciplinary reasoning and understanding (Libarkin & Ording, 2012). Systematic analysis of student writing can establish how well students are learning the ways of knowing in a discipline. The goal of this study was to triangulate data from students’ educational past, also known more broadly as learning analytics, with their current writing performance in a semester‐long course, so as to inform course and curricular innovations that support student learning in STEM disciplines. Specifically, this study examined the relationship between students’ scientific writing skills in an upper‐level undergraduate science course in animal behavior, with their prior performance in college courses, research and/or internship experiences, as determined by course credit, and performance on standardized college admission tests (SAT and/or ACT). Animal Behavior is a 3‐credit lecture course with no pre‐requisites, meeting twice weekly for 75 minutes and typically enrolls junior/senior STEM majors. In the semester studied, 54 students completed the course (54% female and 46% male), and 90% were majoring in the life sciences. A rubric adapted from the Dreyfus model was used to examine students’ abilities to synthesize primary literature, communicate ideas to both a scientific audience and a lay audience, and apply the information to new areas of scientific study (Dreyfus, 2004). Using this rubric, student work was scored by six graduate student and faculty raters from the life sciences and science education. Each assignment was scored by three different raters, with inter‐rater reliability assessed. Information about students’ academic experiences was gathered from the university's assessment and research office. This study was approved by our institutional review board, and all assignments were coded with unique values to maintain student confidentiality. Preliminary analysis found students who completed the university's Biological Mechanisms course tended to synthesize the scientific literature better, and contrary to what was hypothesized, statistics course completion did not enhance students’ abilities to synthesize primary literature when compared with students that did not complete a statistics course. This presentation will demonstrate how triangulating different forms of evidence about student writing can inform our understanding of the knowledge gained and the gaps that persist in student content, process and reasoning in science. Examining student performance in past courses and on standardized college admission tests in the context of current course work can allow us to introduce scaffolding strategies to support conceptual understanding and the ways of knowing inherent to science.