You wish to compute a 95% confidence interval for the population proportion. How large a sample should you draw to ensure that the sample proportion does not deviate from the population proportion by more than 0.09? No prior estimate for the population proportion is available.