Population 1:
p
1
=
n
1
=
Population 2:
p
2
=
n
2
=
5000 samples from each
Reset
mean =
p̂
1
-p̂
2
=
Actual diff of sample props