1.Define the user population and use data obtained from measurements made on that population.
2.Consider factors that might interfere with the assumption of normal distribution of scores. For example, in some countries, stature may be negatively skewed because many individuals do not attain their potential stature because of disease or malnutrition.
3.Remember that many anthropometric variables are measured using seminude subjects. Allowance for clothing is often necessary when designing for real users.