Awesome forum!

My question does not look too complicated, but around my research centre here nobody knows what to do with my problem... There it goes...

I'm collecting 2 different variables during a session of 30 minutes of cycling. One is Force, and the other is an electrical signal that comes from the muscle (EMG). That gives me 2 curves composed by 500 points, over the 30 minutes.

I get these datasets from both legs of 5 subjects, so that I ended up with:

Subject1: Force_Left, EMG_Left ; Force_Right, EMG_Right

Subject2: Force_Left, EMG_Left ; Force_Right, EMG_Right

Subject3: Force_Left, EMG_Left ; Force_Right, EMG_Right

Subject4: Force_Left, EMG_Left ; Force_Right, EMG_Right

Subject5: Force_Left, EMG_Left ; Force_Right, EMG_Right

Both variables (force and EMG) change with time due to muscle fatigue. Normally they decay in a exponential way, but not always. Although, they seem to always change in a synchronised way.

The question I want to answer with my research, is to find out what is the relationship between these 2 variables. In other words, I want to know whether or not the EMG curves reflect (or even predict-next step) the force decay due to fatigue.

I've already thought about averaging legs within subjects, and them average the 5 curves of each variable to compare the final Avg Force with the final Avg EMG... But it just does not feel right, once individual variations (that could be happening in both the EMG and Force curves, so indicating relationships) coud be all masked by doing that...

If anybody has got a solution or even a "jumpstart" is more than welcome to contribute!

Cheers!