Correct

Think about this. A pretest cannot prevent regression
to the mean. All it can do is to establish a measurement
at some level prior to the manipulation. Whatever that
pretest level might be, it might be spuriously high or low
and, if so, any subsequent measures would have a
good chance of reflecting a more normal mean level,
that is, returning (regressing) to the mean.