In: Statistics and Probability
(a)
Randomization is an important feature in a controlled experiment because it has the following advantages ;
(i) Randomization eliminates bias in treatment assignment: in particular, Selection Bias, Allocation Bias and Confounding Bias
(ii) Randomization facilitates blinding of identity of treatments from researchers, participants and assessors, which minimizes Assessment Bias.
(iii) Randomization permits the use of probability theory to express the likelihood that any difference in outcome between treatment groups merely indicates chance.
(b)
Confounding Variable is defined as a variable that influences both the Dependent Variable and Independent Variabe, causing a spurious association between the Dependent Variable and Independent Variabe,
Randomization gives control over confounding variables that cannot be held constant by generating groups that are comparable wit respect to confounding variables.