Indeed. I think it is also important to note that the unknown factor can be something very abstract. For example if you measure the oscillations of two pendulum

Indeed. I think it is also important to note that the unknown factor can be something very abstract. For example if you measure the oscillations of two pendulums with the same properties, they will also be correlated, but simply because they are driven by the same underlying physics...