Year: 2006

Name: Fortus, David

Title of dissertation: “Design-Based Science and the Transfer of Science Knowledge and Real-world Problem-Solving Skills”

Chair of dissertation: Krajcik, Joeseph., Marx, Ronald W.

Institution: University of Michigan

Abstract:

Design-Based Science (DBS) helps students develop new scientific knowledge and problem-solving skills in the context of designing artifacts. This pedagogy was developed as a response to the potential problem of transfer of knowledge from academic settings to extra classroom environments. This dissertation describes DBS in detail and attempts to answer three questions: (1) Do DBS curricula support students' efforts to transfer newly constructed science knowledge and 'designerly' skills (Baynes, 1994) to the solution of new real-world design problems in an extra-classroom setting? (2) Do DBS curricula support students' efforts to construct new scientific knowledge? (3) Do DBS curricula support students' efforts to develop 'designerly' problem-solving skills?

Ninety-two students attending a public high school serving a working class community participated in the consecutive enactments of three different DBS units over one school year. The analysis of pre- and posttests and of artifacts created by the students demonstrated that substantial knowledge was constructed during each of the enactments, with the tests leading to effect sizes of 2.1 on the first unit, 1.9 on the second, and 2.7 on the third.

After each enactment the students solved a new design problem as a transfer task. The transfer tasks were unsequestered, unsupported by the teacher, lasted three days, were done in the school's library, required new learning, and were solved in groups of four. In order to generate an individual measure of transfer, the students responded to an individual post-transfer written test after each transfer task was completed, that assessed their understanding and recollection of the solution their group submitted. For all three units there was a stronger correlation between the individual transfer scores and posttests scores than with pretest scores, indicating that the knowledge and skills that were constructed during the enactments supported the solution of the transfer tasks. The correlations with the posttests increased from one enactment to the next, demonstrating that the students' transfer performance improved as they gained more experience in DBS classrooms.

Potential threats to the study's internal validity that were identified and discussed were improved teacher proficiency, the nature of the transfer tasks, the difficulty of the science content covered by the units, the similarity between the units and the transfer tasks, and the similarity between the transfer tasks.

This dissertation demonstrates that: (a) appropriate learning environments can foster transfer, (b) transfer performance can improve over time, and (c) that it may be necessary to rethink and redefine the procedures for identifying and assessing real-world transfer.