This paper demonstrates the use of principal component analysis to systematically identify aberrant observations in a multi-dimensional context to ensure sample homogeneity. Prior to quantitative analysis based on survey data such as diversified Saskatchewan grain farms, a careful examination of the data using this technique is illustrated.
