Data Research and Development Center and IERI Research Cmmunity header

The IERI Research Community :: Projects

Computational Literacy: A Study of the Efficacy of Computational Science in High School Biology and Earth and Space Science Classrooms

Margaret Honey

Barbara Helland, Susan Ragan, Paul Gray, Cornelia Brunner


Background: As computational models and simulations continue to grow as a means of representing information and conducting research, students need to understand both the power and limitations of these kinds of data representation systems. This means understanding not only how these tools can support research in a wide-range of scientific disciplines, but also the mathematical, computational, and content-based assumptions that underlie computational models and simulations. In order to do this, students today must develop computational literacy--an ability to understand the kinds of questions that computational techniques answer, to determine legitimate relationships among variables, to visualize mathematical functions and scientific models, and to summarize and draw scientifically valid conclusions from data.

Purpose: A multidisciplinary research and development team has developed an approach for teaching computational science that starts with a comprehensive, visual simulation of computational data and encourages students to manipulate the scientific and mathematical elements that factor into the underlying computational model. The overarching objectives of this three-year project are to:

  1. Determine the learning efficacy of the current approach to teaching computational science across two curricular areas--biology and earth/space science--in a range of school and classroom contexts.
  2. Investigate how promising practices, tools, and curricula can scale beyond test bed programs and be disseminated to a broad population of teachers.
  1. Ensure quality professional development resources for teachers that are cost-effective, not dependent on outside experts, and focus on pedagogical strategies that support discovery and exploration rather than transmission of information.

    Intervention: During the first two years (September 2004-2006) of the project, four Internet-accessible, Java-based simulations were developed, covering four core high school science topics: Population Dynamics, Disease Spread, Carbon Cycle, and Rock Cycle. Each interactive, agent-based or equation-based simulation allows students to modify variables or rules in the underlying model. To analyze the results of these modifications, students can either step through the action, run it in its entirety, or replay portions of the simulation to uncover the change between steps. Additionally, the student can view the changes graphically or by reviewing tabular data. Taken together, these options provide abundant opportunities to understand the model underlying the simulation.

    To further support inclusion in the classroom, each simulation is part of a topic module that is housed on a website called "The Computational Laboratory" ( Each module comprises an introductory text that creates a framing narrative for an investigation; the simulation itself, which is contextualized through "scenarios," narrative statements or mini-stories that name and situate a particular instance of the scientific topic under investigation; a set of Driving Questions to enrich and guide students investigations of each scenario; and two diagnostic assessments (a pre/post test and far transfer task). The pre/post test (administered before and after use of the simulation) will indicate whether exposure to computational modeling has improved students' understanding of the scientific phenomena described in the models. The far transfer task will assess the students' conceptual understanding of the dynamic and interactive nature of the phenomena or model. The Computational Laboratory site, which is still under development, is intended to be a virtual science laboratory designed for students and teachers to explore these computer simulations in meaningful ways. When completed, it will disseminate information about computational literacy, the CompLit Project, models and simulations in general, as well as provide related resources and pedagogical tools to support the use of these simulations in the classroom.

    Setting: Participants are being recruited from urban, rural, and suburban high schools in Maryland, Iowa, and Tennessee.

    Research Design: The final year commences September 2006 and will involve the research portion of the project. The overall objective is to investigate whether highly-scaffolded computational simulations enable students from diverse academic, ethnic, and socio-economic backgrounds to perform effectively on the ability to: (1) Describe the affordances and limitations of the model in real-world situations; (2) Solve novel problems related to, but not the same as, the computational problems they have been working on in their courses; (3) Demonstrate a conceptual understanding of the dynamic and interactive nature of the phenomenon or model; and (4) Show progress or proficiency on relevant testing items.

    The team is currently recruiting 72 teachers--36 treatment (18 in earth/space science, 18 in biology) and 36 control. This will yield a sample of roughly 2,160 students (72 classrooms x 30 students per class), distributed between the treatment and control groups. For each simulation, the classroom teacher will be required to spend one 90-minute period using the unit with the students and another 45 minutes (not continuous) to provide a wrap-up discussion and administer the diagnostic assessments. In addition, block randomization will be used when assigning schools, to ensure equitable economic and ethnic representation in the treatment and control categories.

    The outcome analysis will involve examining the data with respect to how students performed on the diagnostic tasks and Driving Questions, as measured by a rubric, and differential performance based on student background variables. We will conduct a regression analysis, using student background characteristics as the independent variables; we also will conduct an analysis using teacher background variables. These will be done with the total sample, and then separately for each content area to determine the extent to which differences across the courses may exist.

    For the far transfer task, we will first examine performance using a two-way ANOVA, with treatment condition as one dimension and general versus honors track as the second dimension. We will then conduct regression analyses, entering student background variables and teacher background characteristics. Again, we will examine differences among the two course content areas.

    Findings: Final project findings will be made available by Fall 2007.

    You can learn more about this project by visiting: