A class exercise was designed for a college-level geochemistry class to promote inquiry and student participation. In this exercise, real soil data available online was analyzed to evaluate geochemical associations among different soil orders and as a screening tool for anthropogenic metal contamination. Students were asked to read a peer-reviewed research article and use the methods in it as a model for analyzing their dataset. The exercise provided a setting to review and reflect on the changes that rocks undergo to produce soils, with ion substitution and formation of clay minerals as key steps in this process. Both active learning and cooperative learning were involved. Students made decisions about which path to take toward reaching a common goal; they first worked independently and later discussed the results as a group, comparing between three statistical methods applied with the data with respect to their advantages and limitations. The real-life dataset exposed students to the common shortcoming of having less-than-ideal coverage of data forced them to make decisions on how to proceed further with their analysis, and provided a good example of how scientific research is conducted. Our approach allowed students to actively engage in their learning and reach their goals, which were evidenced by their comments and their zeal to complete each part of the exercise; however, historic data collection is required to formalize these assertions.