Been musing about student choices within assignments, specifically in my field of geotechnical engineering.  How to do this?

Consider an assignment calculating seepage below a dam or levee resting on layered soil, for example.  The designer must have values of permeability for each soil layer.  Rather than the more traditional approach in which students are given these parameters, why not give the students an abundance of test results for the soils and let them choose the parameters from the results as they would in practice.

The assignment would then be evaluated as follows:

• Say 80% of grade is assigned based on the design analysis (the primary learning objective of the assignment).
• The remaining 20% would be assessed based on HOW the parameters for the analysis were chosen (secondary objective).

Students would be allowed to use all, some, or none of the provided test results to determine the parameters for their analysis.  If few or none of the tests are used, the students must justify this and explain the judgment calls that they have made.  If more or all of the tests are used, the students will see that getting more technical data is “expensive,” and does not necessarily result in more accurate analyses.  This approach would “cost” the students more study hours just like it would cost their clients more in the future.