Correlated Samples
nNatural pairs
¤researcher does not pair
¤naturally occurring pair
¤pairing is based on logical, memberships, etc
nMatched pairs (split-litter for animals, twins, siblings)
¤researcher controls the pairing
¤may pretest to determine pair mate
¤randomly assign each pair mate to treatment & control
nRepeated measure
¤pre and post test design
¤each participant serves as own control
nIf t-tests are significant Ð then difference in participants is due to the treatment