(206a) Student Learning in Virtual Laboratories

In the context of providing an effective capstone experience in experimental process design, with funding from NSF CCLI (proof-of-concept and Phase 2) and the Intel Faculty Fellows Program, we have developed two virtual process laboratories, the Virtual Chemical Vapor Deposition (CVD) laboratory and the Virtual BioReactor laboratory. In a virtual laboratory, simulations based on mathematical models implemented on a computer can replace the physical laboratory. Since real systems do not deterministically adhere to fundamental models, random and systematic process and measurement variation are added to the output. In our case, students engage in the virtual laboratory in a situated manner, where they attempt to optimize the performance of an industrial process by investigating the effect of the input variables. In the School of Chemical, Biological and Environmental Engineering (CBEE) at Oregon State University (OSU), the virtual laboratory capstone project has been used since Winter 2005, where 270 students in teams of 2-3 have completed the assignment. Through the use of cyber-enabled networks and communications technologies, the laboratories are also remotely available to other institutions. The Virtual CVD laboratory project has been completed remotely at the University of Oregon, UC Berkeley, and the University of Montana, as well as at the high school and community college levels. In the latter cases, the instructional design was significantly modified to align with the knowledge and abilities of high school students. In addition to the software development, instructional design and implementation, we have been developing methods to measure the cognitive processes of students undergoing the virtual laboratories, the social interactions between members of the teams, and student expectations of and metacognitions about their virtual laboratory experience. By simulating the physical operation of the process and metrology equipment, this cyber-enabled learning tool greatly simplifies the physical aspect of performing the experiments and allows a student to experience a different emphasis on the experimental design process than he/she typically encounters in a physical laboratory. This feature allows instructors to focus on aspects of student learning that have been historically elusive.

The study reported investigates student cognitions and metacognitions as they engage in the virtual laboratory project. The mixed methodological perspective of this research targets key project activities to collect data. A primary means of data collection is Protocol Analysis where student groups ?think aloud? as they perform the assignment. In addition, samples of written work (Journals and Final Reports) have been analyzed and a method is being developed to characterize student groups' model development. Finally, student metacognitions are characterized from coded student survey responses after completion of three different laboratories, two physical laboratories and a virtual laboratory.

Task analysis of ?think-aloud? sessions has verified that students are engaged in the intended, iterative experimental design approach of practicing engineers. In all cases, students engaged in multiple ?design? and ?analysis? loops. Due to time constraints, this type of iterative experimentation is difficult to achieve in a physical laboratory. Additionally, this laboratory experience was demonstrated to promote higher level cognition in students. Modified Perry's levels were applied to quantify the team's tolerance for ambiguity. Evaluation results elucidating tolerance for ambiguity indicate that by completing this open-ended problem most students evolve past ?blind acceptance of authority? and become aware of a ?multiplicity of views?; however, while some students continued to climb Perry's levels, eventually becoming comfortable with the idea of ?contextual relativism,? other students did not. An interesting parallel to these differences is found in the nature of the sociocognitive interactions found in the different student teams; these interactions seem to be able to either promote the desired learning, or they can be detrimental to the intended learning outcomes. Further research into the impact of reflection and sociocognitive interactions is planned to address these issues more fully.

Preliminary results from this model representation scheme are presented for two industrial scale virtual laboratories, one based on a transient biological system and one based on a steady-state chemical system. Different types of qualitative and quantitative models are evident in the students' solutions and can be generally related to differences in the type of knowledge structures of the physical systems embodied by each of the virtual laboratories. Student groups also show distinct differences in ability to apply schematic and strategic knowledge, and strength in one knowledge type does not necessarily indicate strength in the other. Students' perceptions of the three different laboratory experiences are discussed from the focus of intended (metacognitive questions) and actual learning (cognitive questions). Student perceptions about the laboratory experiences were correlated to student performance in the class, as measured by the students' scores on all graded assessments for the course. Analysis of metacognitive statements of students show enhanced awareness of experimental design, and greater occurrences of critical thinking and higher order cognition in the virtual laboratories.