What is the p-value of 200? Actually Tang, if you really wanted to show that the data wasn't reliable you could have just said that the 200 people that were polled are not random. 200 people can be significant. We don't know. No confidence level or margin of error have been identified.