If an experimenter conducts a t-test for independent means and rejects the null hypothesis, the correct interpretation is that:

* he samples were from populations that were actually dependent rather than independent.

* The variance of one sample is so much larger than the variance of the other sample that the variances of the parent populations must not have been the same after all.

The variance of one sample is so much larger than the variance of the other sample that the variances of the parent populations must not have been the same after all.