Understanding, Evaluating, and Diagnosing Undergraduate Student Difficulties with Graph Choice and Construction

Aakanksha Angra, Purdue University


Creating effective graphical representations of biological data is an essential component in the practices of science and involves engaging concepts and skills of quantitative literacy. With undergraduate biology students increasingly involved in scientific inquiry and experimentation, they are faced with the task of choosing and creating appropriate graphical representations of their data to communicate their findings. However, difficulties with graph choice and construction that were previously documented in literature, still exist today at both the K-12 and undergraduate levels. The purpose of this dissertation is to understand the reasoning involved behind choosing certain graph types and the process that occurs during graph construction, and to design and validate instructional materials to improve graphing skills. The first chapter reviews recent policy documents and relevant literature that have stressed the importance of graphing skill development. Although graphing has been heavily emphasized at the K-12 level and in the context of math and physics, the stepwise thought process and reasoning that determine how the graph is constructed and the final message it conveys are not well understood. In chapter two, I attempt to understand these reasoning that occurs during graph choice and construction by studying expert and novice biologists. Clinical think-aloud interviews were conducted and participants were presented with a small data set and asked to construct a graph using pen and paper. In chapter three, I look at how graphs are constructed in a naturalistic, classroom setting. In Spring 2013 and 2014, students in an upper level physiology laboratory engaged in inquiry-based labs, which required them to work in a team to design experiments, collect data, and present these findings in an oral presentation. Students engaged in guided reflective practices multiple times over the course of the semester, which forced them to evaluate their graph choice and describe the advantages and the disadvantages of their graph. The work described in fourth chapter utilized findings from the second and third chapters, as well as existing literature to develop instructional and learning tools aimed at improving reasoning with graphs. These tools are: the step-by-step guide, guide to data displays, and the graph rubric. The step-by-step guide was informed by the data from the think-aloud interviews (chapter 2) and its purpose is to provide students with a framework for data presentation, as practiced by experts. The purpose of the guide to data displays is to inform students of various types of graphs, their usage, advantages, and disadvantages. The purpose of the graph rubric was to help instructors provide quick and consistent feedback on students’ graphs and for students to use when constructing and critiquing graphs. The graph rubric was informed by: seminal literature in math and science education that informed the 12 assessment categories, expert-novice graphing interviews (chapter 2), and student graphs and reflections (chapter 3). The rubric was validated in three ways: assessing graphs from five introductory biology textbooks, graphs generated in the classroom, and graphs from the science literature. Chapter 5 used the cognitive apprenticeship model and tested the utility of the instructional and learning materials mentioned in chapter 4 in an upper-level physiology laboratory classroom (same setting and curriculum as chapter 3). Data for this chapter were collected during the Spring 2015 and 2016 semesters. Overall findings from this dissertation elucidated the presence of graphing competencies and difficulties in clinical and naturalistic settings in undergraduate biology students, graduate students, and professors, and informed the development and validation of three instructional and learning tools. These materials have the potential to resolve persistent difficulties with graphing and can be incorporated in teacher education and implemented in science classrooms at the undergraduate and K-12 levels.




Gardner, Purdue University.

Subject Area

Biology|Education|Science education|Higher education

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server