AN INTERACTIVE TEACHING DATABASE: HYPOTHESIS TESTING USING THEORETICAL MODELS

Practicing scientists often use theoretical models as a way of performing hypothesis tests on observed data. These hypotheses are usually developed on the basis of ancillary information in the context of a larger system. For example, a practicing hydrogeologist might try fitting aquifer test data with several reasonable hydraulic models before settling on an appropriate interpretation. The choice of "reasonable" models would be based upon the hydrogeologist's understanding of the geology and hydrogeology of the site, and the design of the aquifer test. When we teach aquifer testing in the classroom, however, we might hand the students some hydraulic data and give them one or more models with which to "fit" data. Hypothesis testing in this scenario is impossible, because there is no information on which the student can base an initial guess about outcome of the model. Teaching hypothesis testing in the classroom is time-intensive because both the theory and data must be presented along with site-specific information. The Mirror Lake Interactive Teaching Database attempts to streamline this process by offering lessons on several topics in hydrology using data from the same field site, so that the same background can be used for multiple lessons. Atmospheric, surface-water, and ground-water data are derived from the Hubbard Brook Experimental Forest Long Term Ecological Research Site. Computational models are kept as simple as possible, so that students can concentrate on hypothesis testing and interpretation. For simple models, calculations are made using interactive java/perl scripts. More complex models, such as numerical ground-water flow models, are pre-calculated and then retrieved on-the-fly by the student. Example lessons will be displayed along with general information about the approach.