In one sample t-test, you compare it to some value that you already know to represent the population mean. In the t-test for correlated groups, the statistic is the difference between your two samples, which you test against the population mean difference.