Graeco-Latin square experimental design

An experimental design that allows the researcher to control statistically for three non-interacting

extraneous variables in addition to the independent variable.