I put yes, as surely you had to use it, as in the qustion it did not state that the sample was assumed to be normally distributed, and in order to calculate a confidence intervals the sample means must be assumed to be normally distributed, and the only way you could tell this was by the Cental Limit Theorem as n>30