ABSTRACT
Traditional tests of infant mental development have no predictive validity during the first 1½ years of life. Indeed, predictive validity is so low that actually testing an infant yields virtually nothing that cannot be predicted from parental education or socioeconomic status alone, which yield correlation coefficients of .40 to .60 (McCall, 1979, p. 714). Many interpretations of this shortcoming have been considered (Brooks & Weinraub, 1976, pp. 19–58; Lewis & McGurk, 1972; McCall, 1979; Stott & Ball, 1965; Wilson, 1972), but the generally accepted view is that development occurs in spurts and lags. Not only are delays naturally occurring and a nemesis to predictability, but they are self-correcting. A contrasting account of this poor predictability is proposed here. It is suggested that the measurement of mental ability and the development of expressive capabilities are confounded in traditional tests—a view that has the demonstrable advantage of internal consistency and experimental validity.
