The t test for nonindependent samples is when
◦ matching has been done or in a pre-post design.
◦ the two groups are formed by preexisting samples.
◦ the two groups tested are initially the same on the dependent variable.
◦ the two groups tested are randomly formed samples.