Actually, statistically, about 1,000 people is enough if a truly random sample is used. But if the sample is manipulated, you could make the poll say whatever you wanted.
I think they believe 1300 is needed for the +/- 3%.
1,000 responders taken from 5-7pm on Friday is not very valid.
Not if the sample is under 10% of the population being studied. There is no way that .000004% is a representative sample.
What do you do about people like me who give false answers to polls?
I got called before the 2004 election and said I supported Kerry when the reality is I voted for Bush.