Temperament questionnaires in clinical research.

Parental questionnaires concerning the nine temperament variables studied in the New York longitudinal study have been developed for the measurement of children's temperament at six, 12 and 24 months of age. Retest reliability for the questionnaires was found to be generally satisfactory and at a level similar to that found in other temperament studies. The temperament characteristics of a large standardization sample, studied longitudinally over the three ages, showed little relation to the children's gender, birth order or social class. Factor analyses of the variables showed that two of the patterns were identical at 12 and 24 months of age: one pattern consisted of mood, approach, adaptability and distractibility, and the other of activity, approach and intensity. The standardization sample showed about the same rates of the 'difficult' and the 'slow-to-warm-up' temperament types, as found in early samples. The stability of the subjects' temperament scores over time was low to moderate, and a majority of the subjects changed their temperament type over the 18 months covered by the longitudinal study. The use of temperament questionnaires in clinical research is illustrated in three current studies.