Children as independent variables: Some clinical implications of child-effects