Some assumptions and difficulties with common statistics for sensory analysis See original record