Statistical Consultancy, Modelling & Data Analysis
Imagine a scenario where consumers are given 5 products to taste; or employees in a B2B (Business to Business) project are given 5 concepts to evaluate. Many researchers do not understand that order and adjacency can dramatically affect the outcome; and they respond to the problem by random allocation of samples not realizing that this can lead to imbalance. We would respond by using a mutually orthogonal Latin square design thus ensuring that carry over effects are balanced and minimized.
Banner headings for tables are often under-utilized for interpretive purposes, careful examination and specification of breaks can add considerable insights into interpretation and presentation.