S. Phillips, M. Dud\'ık, and R. Schapire. ICML '04: Proceedings of the twenty-first international conference on Machine learning, page 83. New York, NY, USA, ACM, (2004)
D. Scott, S. Al-Jibouri, M. Mawdesley, and G. Long. International Technology, Education and Development Conference (INTED 2009), March 9-11, 2009, Valencia, Spain, (2009)
M. Feuer, L. Towne, and R. Shavelson. Educational Researcher, 31 (8):
4(2002)“For example, when well-specified causal hypotheses can be formulated and randomization to treatment and control conditions is ethical and feasible, a randomized experiment is the best method for estimating effects” p. 8.