Definition:
- For any two random variables and , which will make up a new random variable of different distribution with distinct observations consists of both and
- Unlike related variable, where observations of the same distribution is added
- For both discrete and continuous