Being careful with PASCO's kinetic friction experiment: Uncovering pre-sliding displacement?

The widely used PASCO laboratory equipment is an excellent way to introduce students to many topics in physics. In one case, PASCO's equipment may be too good! Various experiments exist for calculating the kinetic coefficient of friction by measuring the acceleration of a sliding object under some constant force. With ever more accurate equipment, such as electronic motion sensors, students are capable of measuring motion over quite small time intervals. In measuring motion including friction, PASCO equipment can record more complicated aspects of friction associated with the transition between static friction and kinetic friction. This serves as an excellent exercise to introduce some fine details of friction not typically discussed in an introductory physics course. In fact, if one does not consider these fine details (or at least does not omit them) the "canned" PASCO friction lab will yield wildly spurious results. The erroneous results obtained are due to a nonconstant "recoil" acceleration during the first ~0.2 seconds of motion. The problem does not show up in the PASCO instructor's manual because the manual restricts the experiment to a small range of low applied forces for which the effect is minor. The recoil was actually observed previously¹ but was written off as equipment noise. If one ignores this "noise," a relatively constant coefficient of kinetic friction can be found in this lab experiment, as we will show. We describe here the original experiment, a second experiment to rule out that this is an experiment-specific phenomenon, and how the experiments can be used for two or three different topics. Finally, we tabulate results and discuss what may be causing this "recoil."