We present results from an experiment in which 55 human subjects each interact with a dynamic system 40 times over a one-week period. The subjects are divided into 5 groups of 11 subjects. Each group interacts with a different system. We use the experimental results to determine characteristics (e.g., system zeros, relative degree, phase lag) that can make systems difficult for a human to control. We use subsystem identification to model the control strategies (feedback and feedforward) that each subject uses during their first and last trials. For each group, the average identified feedforward controller on the last trial approximates the inverse dynamics of the system with which the subjects in that group interacted.