Find a poll reported in the news and write an explanatory article about the results of the poll. Explain to the reader how you can use the normal distribution to compute the likelihood that the sample proportion (for a randomly chosen sample) is in a given range as long as the sample is not too small (the sample must be big enough that the sampling distribution is about normal--explain how big the sample must be for this to occur), what this range, called the margin of error, is and what it means. (For example, explain that it does not give a guarantee that the actual population proportion is within the margin of error but that it instead means that if the poll were run a bunch of times that 19 times out of every 20 times that the poll were to be run the population proportion would be expected to be within the range given by the sample proportion, plus or minus the margin of error.) Also explain how the sample size affects the margin of error and why the same sample size gives the same margin of error whether the population being sampled is small (relatively speaking, such as occurs for a state poll), or huge (such as occurs for national polls). Point out that because of this fact, one can obtain reliable measurements of national sentiment by polling what may seem a surprisingly small sample of a thousand or two.