A multivariate analysis which maximizes the spread of data by plotting covariance values on sets of axes in multidimensional space allowing correlations which may have been hidden in the data to be identified. The first principal component corresponds to the first axis in multidimensional space and describes the majority of the spread of the data, subsequent higher order principal component axes are orthogonal to the first axis. Higher order axes display progressively less variation, where the data is less correlated and more representative of statistical noise.