The validity of psychiatric diagnoses: The case of ‘specific’ developmental disorders

We tested whether developmental coordination disorder (DCD) and mixed receptive expressive language disorder (RELD) are valid diagnoses by assessing whether they are separated from each other, from other childhood disorders, and from normality by natural boundaries termed zones of rarity. Standardized measures of intelligence, language, motor skills, social cognition, and executive functioning were administered to children with DCD (n = 22), RELD (n = 30), autistic disorder (n = 30), mental retardation (n = 24), attention deficit/hyperactivity disorder (n = 53) and to a representative sample of children (n = 449). Discriminant function scores were used to test whether there were zones of rarity between the DCD, RELD, and other groups. DCD and RELD were reliably distinguishable only from the mental retardation group. Cluster and latent class analyses both resulted in only two clusters or classes being identified, one consisting mainly of typical children and the other of children with a disorder. Fifty percent of children in the DCD group and 20% in the RELD group were clustered with typical children. There was no evidence of zones of rarity between disorders. Rather, with the exception of mental retardation, the results imply there are no natural boundaries between disorders or between disorders and normality.

► Children with developmental disabilities are typically misclassified based on ability scores.
► The most common classification error is for disordered children to be classified as typically developing.
► Of the disorders examined, only mental retardation was separated by a natural boundary from other disorders.
► Motor skills and communication disorders are separate from each other, but not from autism or typical development.
► Cluster and latent class analyses distinguish higher from lower functioning children, not distinct disorders.