A 90% confidence interval (CI) is created to estimate the proportion of television viewers in a certain area who favor moving the broadcast of late weeknight news one hour earlier. This CI is created using a simple random sample of 9,000 viewers in the area. Assuming that the sample proportion does not change, what would be the relationship between the width of this CI and the width of a second 90% CI that is created based on a sample of only 1,000 viewers in the area