This paper describes the use of the computer‐based interview as a research technique for investigating how students think about physics. Two computer programs provide the context: one intended for instruction, the other for research. The one designed for use as an instructional aid displays the motion of a ball rolling along a track that has level and inclined segments. The associated motion graphs are also shown. The other program, which was expressly designed for use in research, is based on the simulated motion of a modified Atwood’s machine. The programs require students to predict the effect of the initial conditions and system parameters on the motion or on a graph of the motion. The motion that would actually occur is then displayed. The investigation focuses on the reasoning used by the students as they try to resolve discrepancies between their predictions and observations.