ABSTRACT
Researchers interested in a particular syndrome or developmental disorder often wish to elucidate the cognitive and language characteristics associated with that group, with the goal to determine characteristics that are universal within and specific to the target group. To this end, researchers typically compare a group with the target disorder to contrast groups matched for certain variables (CA, IQ, and MA are common choices). If the target group performs significantly worse than both normally developing and delayed contrast groups on the dependent variable(s), it is commonly concluded that the target group evidences a universal deficit on the dependent variable and that the deficit is specific to that group, For example, suppose that a group of children with autism responds significantly less often to an experimenter’s bids for joint attention than either a languagematched group of children with other forms of developmental delay or a language-matched group of children who are developing normally. Given this pattern of results, researchers typically would conclude (as Sigman & Ruskin [1999] did) that deficits in responding to bids for joint attention were universal and specific to children with autism. Sometimes such conclusions are based on comparisons of only two syndrome groups. As another example, suppose a group of adolescents with Williams syndrome performed very well on a series of syntactic measures and significantly better than a CA-and IQ-matched group of adolescents with Down syndrome. Given this pattern of results, the conclusion drawn by Bellugi and her colleagues (e.g., Bellugi, Marks, Bihrle, & Sabo, 1988; Bellugi, Wang, & Jernigan, 1994) as well as many other language researchers (e.g., Jackendoff, 1994; Pinker, 1999) was that the language ability of individuals with Williams syndrome was both much higher than expected for their level of mental retardation and essentially normal.
