​for a repeated-measures design, the differences between treatments may be caused by a treatment effect or by sampling error (random factors), but they cannot be caused by individual differences.