How to use Central Limit Theorem here?

Suppose you'd lke to estimate the procentage of children (<18) 
in a population by picking out a person and then putting him/her
back. How many times do we need to make a "pull" if we'd like
to be 95% sure that our percentage deviates at most 2% from 
the atuall one?

Solution:
I use X~Ber(p) and create
( Sum[ X_i, i=1, i=n ] ) / n  -  n*p
since p is the E[X]. What i need int order to use CLT is the 
variance but i can't compute it (or rather, i can but it will be
expressed as p-p^2 which make the final answer a function
of the variable p). Any suggestions to a solution?

-- 

Kindly
Konrad
---------------------------------------------------
May all spammers die an agonizing death; have no burial places; 
their souls be chased by demons in Gehenna from one room to 
another for all eternity and more.

Sleep - thing used by ineffective people
            as a substitute for coffee

Ambition - a poor excuse for not having
                 enough sense to be lazy
---------------------------------------------------