Which of the following conditions is/are necessary for a hypothesis test concerning inference on two-proportions?

I.The data must be from a simple random sample.

II.${ { n }_{ 1 } }(p),{ { n }_{ 1 } } \times (1-p),{ { n }_{ 2 } }(p)$ and ${ { n }_{ 2 } }(1-p)$ must all be greater than $5$, where ${ { n }_{ 1 } }$ and ${ { n }_{ 2 } }$ are the sizes of the first and second samples, respectively, and p is the pooled sample proportion.

III.The samples must be stratified.