A store emails all of its (several thousands of) customers an online survey. Of all its customers, 302 reply to the survey and of these, 215 gave a five-star rating to their experience shopping at the store. The manager could estimate that 215/302 (or about 71%) of all customers would give a five-star rating, but why is a Z-interval not valid for a confidence interval here? Select all that apply. The proportion of the population of all customers is not approximately normal. The number of customers in the sample is not large enough. The sample was not randomly selected. The proportion of the population of all customers is known already.