Computer-based instruction and testing: Preliminary data on student scientific literacy and academic behaviors

Background/Question/Methods

As instructional technology transforms biology instruction it also offers new opportunities to assess student academic behaviors. For instance, the higher-order thinking and procedural skills necessary for success in the sciences are rarely assessed through current paper-based tests, while data on student behavior out of the classroom is difficult to collect. Computer-based instructional tools promise to increase the efficiency with which we can assess student understanding and academic behaviors. In one on-going ecology education project, we are developing a tool to assess student understanding of systems and experimental design, as well as skills in interpreting and creating graphs, using a computer-based test built around an agro-ecological scenario. As part of this study, we collected preliminary data from high school aged students on their confusions in these areas. In a separate study, we are developing tools to track student use of instruction and feedback tools in a digital textbook. We have collected preliminary data on student use of chapter questions with embedded feedback, as well as tracked their use of chapter hyperlinks to primary literature, definitions of terms, and explanations of equations.

Results/Conclusions

In the first study, we found that students commonly valued inputs over outputs in determining population size, failed to construct experiments that would test for multi-variable causality, and had a number of difficulties in representing tabular or graphical data in a graph. These difficulties can be scored automatically by the computer-based test.  In the second study, we found that although students rarely followed hyperlinks to references in the chapters, heavy use of specific links suggest student difficulty with some ecological terms and equations. In addition, students often used question feedback to alter their responses, but this use was inconsistent across questions. Although preliminary, the results from these studies highlight the potential of computer-based instructional and testing tools to improve biology education.