Average Faces Are Average Faces
Photographs of faces of young adult male and female Scots were measured on nineteen frontal dimensions. Measures in each dimension were converted to z-scores and summed for each face. In 1990, J. K. Langlois and L. A. Roggman described the first systematic study to support the claim that faces representing the average of the population are consistently judged as attractive. They digitised photographs of student faces and arithmetically averaged groups of the resulting matrices of numeric grey values to construct composite portraits. They found that attractive composites could be made more attractive by warping shape differences away from the sample mean. Empirical evidence both for and against the “attractiveness is averageness” hypothesis has come entirely from studies using artificially contrived faces. The extent to which these processes produce “average” faces is not always clear. Subjects for the attractiveness judgements were twenty male and twenty female volunteer white New Zealand undergraduate psychology students in their late teens or early twenties.