When an experimenter wishes to ensure equal representation of a highly correlated variable with dependent variable at all levels of independent variable, he/she should use __________ design.
1
Repeated measures
2
Randomized groups
3
Latin Square
4
Randomized block