the property that if we repeatedly selected our random sample in exactly the same way, each time constructing a different 90% confidence interval for \mu, then in the long run 90% of those intervals would contain \mu.

the property that if we repeatedly selected our random sample in exactly the same way, each time constructing a different 90% confidence interval for \mu, then in the long run, these intervals would overlap and the overlapping interval would converge to the actual 90% confidence interval for \mu .

the property that if we repeatedly selected our random sample in exactly the same way, each time calculating a different sample mean, then in the long run 90% of those sample means will be contained in this confidence interval.

the property that the confidence interval for the population mean \mu calculated from this sample is correct with 90% confidence.