Twins are particularly useful for developmental behavior genetics, if adequate consideration is given to the possibilities of twin-singleton differences and of data from twins being unrepresentative of the entire population. As an alternative to existing analyses in the developmental behavior genetics of cognition, the covariance structure analysis method in biometrical genetics involving twins and other relatives has the potential for dissecting the determinants of the structure of cognitive abilities and how these change during childhood, as well as checking on the adequacy of twin data. Designed with these aims in mind, the La Trobe Twin Study is a mixed longitudinal analysis of 1,356 twins, their siblings, and cousins, who are followed between the ages of 3 and 15 years on a battery of physical and behavioral tests. The criteria for choosing behavioral tests in such a project are discussed, along with the comprehensive series of questionnaires necessary to examine the children's development in the context of the stresses a multiple birth may impose on a family. In discussion of methods of analysis, 2 points are emphasized: (1) Twin-singleton differences are specific to particular abilities rather than being general, and (2) the incidence of family changes and stresses in such that these should be considered among the environmental effects on development along with the more usual indices of family environment.

In 1975 the Australian Council of Educational Research (ACER) conducted a nationwide survey of literacy and numeracy in 10- to 14-year olds. A total of 297 of the 12875 children involved were twins. By age 14, only 42% of the twin boys achieved adequate standards of literacy compared with 71% of single-born boys. The deficit in twin girls was much less and twins of both sexes were only moderately behind in numeracy. A survey of 9-13-year-old twin boys in the La Trobe Twin Study (LTS) produced similar results with 75% being below average in reading skills and 23% behind by 18 months or more, despite above average IQs. The ACER data are corroborated by teachers' reports obtained in the same survey, which indicate also how few of the twins with problems are receiving remediation and the high incidence of classroom problems in spelling and reading reversals. The pattern of mistakes twins make on specific items in the ACER survey can be explained as resulting either from specific cognitive deficits or from problems in concentration. The same factors influence performance on different tasks, so that literacy and numeracy are much more closely interrelated in twins than in singletons, and also correlate more with a measure of verbal intelligence. Implications for genetic analysis of scholastic achievement are examined, centering around the different factor structure of abilities in twins and common family environmental effects which are unique to twins.

In the La Trobe Twin Study, data are collected on temperament and social development in 3-15 year old twins and singletons from four different sources: questionnaires to parents covering development from birth to the present plus the Bristol Social Adjustment Guide completed by the teacher, Sattler's Behavior and Attitude Checklist completed by the tester, and the Qualitative Score on the Porteus Maze Test. Particular stresses are identified which the parents perceive as distinguishing a multiple from a singleton birth. Whereas they perceive no differences between the first and second-born in birth complications, the second-born is judged less favourably particularly in MZ pairs. The distinction continues in the later assessments by the teacher and tester, where in addition the male twins are seen as being different from other children both in cognition and in temperament. It is proposed that social and cognitive development of twins are interrelated and have two unique components, one related to the greater problems accompanying a multiple birth and the other to comparisons between cotwins.

The birth of twins is often accompanied by complications which may affect both twins equally (eg, prematurity) or which may establish differences between the twins (eg, one coming home from hospital first). Parents' ratings of behaviour may reflect and even perpetuate the influence of such variables and it is of interest to see if people less familiar with the twins' history can observe any long-term effects. In the La Trobe Twin Study teachers routinely complete the Bristol Social Adjustment Guide, an assessment of social maladjustment in children aged 5-16 years. Extreme prematurity and problems at delivery were generally associated with the Underreaction syndrome - the child who is unforthcoming, withdrawn and depressed. Withdrawal, depression and maladaptive behaviour to classroom peers were much more common among those twins discharged from hospital after the cotwin. A major mediating factor was that the parents admitted to greatly preferring the one to come home first. These raise questions both about the practice of bringing twins home separately (which occurred in 21% of the cases) and about intervening variables in studies of personality and temperament.

Twins are particularly useful for developmental behavior genetics, if adequate consideration is given to the possibilities of twin-singleton differences and of data from twins being unrepresentative of the entire population. As an alternative to existing analyses in the developmental behavior genetics of cognition, the covariance structure analysis method in biometrical genetics involving twins and other relatives has the potential for dissecting the determinants of the structure of cognitive abilities and how these change during childhood, as well as checking on the adequacy of twin data. Designed with these aims in mind, the La Trobe Twin Study is a mixed longitudinal analysis of 1,356 twins, their siblings, and cousins, who are followed between the ages of 3 and 15 years on a battery of physical and behavioral tests. The criteria for choosing behavioral tests in such a project are discussed, along with the comprehensive series of questionnaires necessary to examine the children's development in the context of the stresses a multiple birth may impose on a family. In discussion of methods of analysis, 2 points are emphasized: (1) Twin-singleton differences are specific to particular abilities rather than being general, and (2) the incidence of family changes and stresses in such that these should be considered among the environmental effects on development along with the more usual indices of family environment.