Is there a relationship between children's mental disorders and their ideal family constructs?

Psychiatric outpatients (N = 72) and nonclinical children (N = 72) from 1st to 12th grades portrayed their families as they wished they were (ideal representation), using the Family System Test (FAST), a figure placement technique that focuses on cohesion and hierarchy. Results showed that constructs of ideal family relations were related to age group (1st to 3rd graders versus 4th to 12th graders) and to clinical status (clinical versus nonclinical) but not to the type of mental disorder. First, younger children portrayed less cohesion than did older ones. Second, clinical respondents were less likely to represent cohesive family patterns and were more likely to report ideal family situations as special events. Results are discussed from a developmental as well as from a family systems perspective.