The Paired-Samples T-Test is used to find out whether two related measurements from the same group or units are different or not. These two measurements typically represent different times (e.g. a pre- and post measurement with a treatment between them) or two different but related conditions. The Paired-Samples T-Test needs two related continuous variables that both are measured on the exact same scale. The main assumption to be tested statistically is that the paired differences should follow an (approximate) normal distribution.