To Construct a Confidence Interval for the Difference of Two $5 \%$

To construct a confidence interval for the difference of two population proportions the samples must be independently obtained random samples, both must consist of less than $5 \%$ of the population, and

A) both np1 $( 1 - \hat { p } 1 ) \geq 10$ and $n \mathrm { p } 2 \left( 1 - \hat { p } _ { 2 } \right) \geq 10$ must be true.
B) only one of $\hat { \mathrm { np } 1 } ( 1 - \hat { \mathrm { p } } 1 ) \geq 10$ or $\hat { n p } _ { 2 } ( 1 - \hat { \mathrm { p } } 2 ) \geq 10$ must be true,
C) $\hat { n p 1 } ( 1 - \hat { p } 1 ) + \hat { n p } _ { 2 } ( 1 - \hat { p } 2 ) \geq 20$ .
D) $\mathrm { np } 1 \left( 1 - \hat { p } _ { 1 } \right) \mathrm { np } 2 \left( 1 - \hat { p } _ { 2 } \right) \geq 100$ . 4 Test hypotheses regarding two proportions from dependent samples.