Independent random samples from two normally distributed populations give the following results:
nx = 15 x-bar = 400 sx = 10
ny = 13 y-bar = 360 sy = 40
If we do not assume that the unknown population variances are equal, what is the 90% confidence interval for the difference between population means?