In a block design, why is 'blocking' used?
To remove the need for random assignment
To provide a larger sample size
To reduce variability by grouping similar participants
To increase the number of treatment groups