You're correct. It isn't different. Simpsons paradox is actually a key indicator of a confounding variable.