ABSTRACT
The previous chapter included a review of current conclusions, extracted from correlations among ability tests, about a suitable conceptual structure for abilities. In the history of psychology, studies of intercorrelations were far outnumbered by studies of the predictive power of ability tests, and we move now to a summary of current conclusions on prediction. The conclusions were largely consolidated before Snow entered the field, and he rarely did such research. Snow did much, however, to locate evidence that affcon variables have causal and predictive power, as the term aptitude implies. Research in that vein has led only to scattered (but suggestive) conclusions; this chapter presents some current ideas about practical prediction. The sophisticated prediction studies to which Snow devoted nearly 20 years are the final topic of the chapter. l 4 Aptitude-treatment interaction (ATI) studies in education hypothesize that a student is likely to have greater readiness for one type of instruction than another; recall the example in Exhibit Psc. The extensive empirical research on AT1 has raised many conceptual questions, but conclusions from statistical comparisons across experimental treatments are, by themselves, too specific to generate theory. The last parts of the chapter illustrate the style and rationale of AT1 studies Snow directed; these laid the ground for the studies of process to which he turned following the appearance in 1977 of the Cronbach-Snow book on ATI.
