Monday, March 2, 2009
experimental design for K-zoo verification
An interesting discussion has been taking place between Zach and Rob regarding the best way to run the comparison between the WiiCane and the Optotrak systems that we have planned for early June. As everyone already knows, our project calls for running test subjects along a 30 foot straight line course using both systems. both apparatus call for mounting IR light sources on the cane, and Zach and Rob think they might interfere with one another, which would significantly complicate matters. They also wonder whether it is necessary or important, from an experimental design perspective, to do them simultaneosly, so that we end up with head-to-head, swing-by-swing comparisons for individual running of the course by study participants. I have always thought that they would be done separately, but I can see very good arguments for doing them together. We will be much more confident in the WiiCane system's accuracy if we can compare results collected simultaneously, because we will see, for each participant, whether the device is accurate or not, and if it is not, we will immediately know which aspects of our system need improvement. But if we do them separately, we will have to aggregate the results and try to draw conclusions, which is certainly possible, but will require more number crunching, and will inevitably be less directly meaningful. I think that Annette, as the Project Evaluator, should weigh in here. Of course, we will not know if this is a problem until we actually try both systems in the presence of IR light sources. I think we need to do this test right away, and certainly not wait until we are in K-zoo to do it. It would be best if we could use a single IR source mounted on the cane that can work for both WiiCane and Optotrak. Having two many devices and cables mounted to the cane and cables will affect weight and balance, and may have other unexpected consequences.