Toward a validation of infant testing.

Abstract

NFANT testing had its origins in the work of Gesell and Cattell some I twenty years ago. Operating on the assumption that the rate and limits of futurd mental development are foreshadowed by the manner and fullness AID Pracntcd at the 1954 Annud Meeting.

Effect of maternal education and ethnic background on infant development.

“…But her study was criticized on the grounds of being confined to institutionalized children. When DQ studies were performed with broader groupings (Illingworth, 1960;Simon and Bass, 1956;Knobloch and Pasamanick, 1963), a correlation was traceable between DQ scores below one year and later DQ and IQ measurements. In this particular study it was seen that the infants of the mothers of high education scored significantly higher in all four areas of development but least in the motor area.…”

Some Problems of Studies Concerned With Evaluating the Predictive Validity of Infant Tests

“…and follow-up test scores. Recently it has been claimed (Simon and Bass, 1956;MacRae, 1955;and Escalona and Moriarty, 1961) that such correlational procedures can be misleading and necessarily restrictive. What is important, it is argued, are not the actual test-score correlations that exist but rather whether, for example, a child classified as "dull" (representing a broad range of possible test scores) retains or changes his classification at follow-up.…”

“…One problem is statistical. For example, Simon and Bass (1956) report as a measure of association contingency coefficients derived from x'^. Their coefficients are undoubtedly in error.…”

“…More recently (e.g. Simon and Bass, 1956;and Knobloch and Pasamanick, 1966) investigators have quite appropriately questioned the conclusions arrived at by many ofthe early researchers. For example, most early studies make use of normal or superior children while several recent studies of the last decade have attempted to show the diagnostic or predictive usefulness of infant tests if defective or retarded children are included in the sample.…”

“…The popular retrospective methodology in which researchers search files for children who, when tested as infants, met certain characteristics is particularly vulnerable on this point. Simon and Bass (1956), in a frequently cited study, attempted to show the predictive validity of infant tests by examining files of 3000 children tested when they were infants. Their samples, representing children selected partly on the basis of age at testing, averaged about fifty children.…”