Research on thinking ordinarily examines individual or group differences in performance under "typical" conditions, although sometimes it attempts to optimize performance by providing hints or special instruction. It often looks at changes in performance over time. A major stream of research on thinking that was begun by Galton studies the intercorrelations among "factors" assumed to reflect stable (and perhaps inherent) mental characteristics (see Carroll, 1978, for a history of mental ability testing and test theory). A positive aspect of much of this research has been its willingness to examine naturally occurring relationships as a source of hypotheses to be tested. But a negative aspect, only partly redressed by the influence of Vygotsky and Piaget, has been the heavy use of elaborate Statistical procedures—such as correlation analysis, regression analysis, and factor analysis—which ordinarily require the questionable assumptions that relationships are linear and effects additive (Hamilton, 1980). Until recently. furthermore, this research has virtually ignored social and cultural influences on Thinking.