According to your findings, I do believe you must be using a high-wing trainer for your experiments. Different aircraft can provide some strange reactions to rudder, especially at a cruise or higher speed. Some machines will actually yaw in the direction of applied rudder, then roll in the opposite direction. The old Dave Platt designed original Contender, by the original Top Flight, was notorious for this. Sterling also produced a large .60 job that was the same, and I designed a sport model along the same lines that also had the characteristic. You are doing the right thing by trying things.

Highwing trainer usually roll well with rudder. (After all, that's how 3 channel works.) I have my students slow fly just near a stall using rudder to maintain level wings and also make turns. They get a better understanding of just how important rudder is. Larger models can be flown to better approaches if you can coordinate rudder with the ailerons, and during the round-out phase use ONLY rudder, rather than aileron unless cross-wind demands cross control touchdowns.

When you are just boreing some holes in the sky, try to control your roll/bank with the rudder and observe the reaction. That will tell you how to use the rudder for any trim needed there and not be doing it backwards. BTDT. [X(]

Make a mid-bank slightly climbing turn, where you can fairly well see the model in plan-form. If the nose appears to be pointing in a direction other than the turn circumference, apply rudder and observe the reaction. You can then notice if the turn is "uncoordinated" and that too will tell you if you need rudder trim, which may then require some other aileron trim. However don't get confused with drift-angle in a wind condition.

Some low-wing models will simply YAW when you apply rudder. Those are the easy ones. Good luck.