The reason we use a z-score from a normal distri- bution in constructing a large-sample confidence interval for a proportion is that

(a) For large random samples the sampling dis- tribution of the sample proportion is approxi- mately normal.

(b) The population distribution is normal.

(c) For large random samples the data distribu- tion is approximately normal.

(d) If in doubt about the population distribution, it's safest to assume that it is the normal dis- tribution.