Calculating a sample size: Each student tosses a coin 10 times and calculates the proportion of heads observed. Use that value of to estimate the sample size needed (call it n1) so that a 90% confidence interval for p will have a margin of error of 0.05. Make additional tosses Page 430 to bring the total to n1, and compute a 90% confidence interval for p.

What proportion of confidence intervals cover the true value of 0.5? Is it close to 90%? If someone’s confidence interval doesn’t contain the value 0.5, should we conclude that the person’s coin is unfair? Or is there a better explanation?