This experiment tests the conservation of momentum on an air track: you launch glider 1 at a stationary glider 2, and light gates read the speeds before and after. The ratio of momentum after to before should be exactly 1. It will not be.
What makes this one different
In viscosity by Stokes’ law the formula was being used outside its domain. Here the formula is exactly right and no instrument is broken — the track is not level, by less than a tenth of a degree. A few millimetres of height difference is enough to push your answer about 5% away from 1.
The teaching point is that you can remove this error without ever knowing what it is. It belongs to the apparatus, not to the physics, so reversing the launch direction flips its sign: launching downhill reads momentum high, launching uphill reads it low. Repeat each setup both ways and average. The trick is called the method of reversal.
It comes with a condition that is easy to miss: reversal only cancels the error when both directions were taken with an identical setup — same masses, same kind of collision, same speed. Try a heavy slow glider one way and a light fast one the other, and averaging leaves most of the error behind while you believe you have removed it. The page counts matched pairs for you and warns when a run has no partner.
A second lesson comes free: in the inelastic collisions the kinetic energy really does drop, by exactly m₁/(m₁+m₂). That is not an error, it is physics. Telling “my apparatus lied” apart from “nature is like this” is the hardest part of lab work.