ABSTRACT
Temperament has been put forward as a major influence on children's vulnerability to a broad range of stresses and therefore of central importance in understanding the development of psychopathology. The sample of twins was divided into three age ranges for purposes of data analysis: 0-2 years, 2-5 years and over 5 years of age. The complete set of ratings obtained from the families of twins provides the opportunity of establishing in a variety of ways the reliability of ratings of temperament. By obtaining independent ratings of the twins' temperament from mothers and fathers it is possible to derive three separate estimates of within-pair twin similarity. The advantage of the twin similarity correlation obtained from different parents rating the two twins is that it overcomes an often-made criticism of the twin method, that is, that greater similarities are obtained for monozygotic twins through biases in parental ratings.
