A subsequent study attempted to relate performance at the racetrack to making stock-market predictions in which the same algorithm was involved. Ceci and Ruiz (1991) asked racetrack handicappers to solve a stock-market-prediction task that was structured similarly to the racetrack problem. After 611 trials on the stock-market task, the handicappers performed no better than chance, and there was no difference in performance as a function of overall cognitive ability. Ceci and Roazzi (1994) attribute this lack of transfer to the low correlation between performance on problems and their isomorphs. "Problem isomorphs" refer to two or more problems that involve the same cognitive processes but that use different terminology or take place in different contexts.

The same principle that applies to adults appears also to apply to children. Carraher, Carraher, and Schliemann (1985) studied Brazilian children who, for economic reasons, often worked as street vendors (see also Nuñes, 1994). Most of these children had very little formal schooling. Carraher et al. compared the performance of these children on mathematical problems that were embedded in a real-life situation (i.e., vending) to problems presented in an academic context (e.g., 2 + 4 = ?). The children correctly solved significantly more questions that related to vending than they did math problems that were academic in nature. When the academic problems were presented as word problems (e.g., If an orange costs 76 cruzeiros and a passion fruit cost 50, how much do the two cost together?), the rate of correct responses was substantially better, but still not as high as when the problems were presented in the context of vending.

This lack of transfer also appears to work in the reverse direction. For example, Perret-Clermont (1980) found that many school children had no problem solving paper-and-pencil arithmetic questions, but could not solve the same type of problem in a different context (e.g., counting bunches of flowers). That is, school children may fail to transfer the academic knowledge to everyday problems.

Roazzi (1987) found similar results when comparing street-vendor children to middle-class school children. He compared the performance of children on a class-inclusion task. To assess the performance of the street-vendor children, the researcher posed as a customer and asked questions about the items to find out if the children understood the relationship among classes and subclasses of food (e.g., mint and strawberry chewing gum as part of the class "chewing gum"). At a later time the same children were given a formal test with the same logical structure, but that was irrelevant to their street-vending jobs. The middle-class children were given the same two tests. Street-vendor children performed significantly better on the class-inclusion task in the natural than in the formal context, whereas middle-class children were more successful on the formal version of the task.

Additional research has shown that the use of complex reasoning strategies does not necessarily correlate with overall cognitive ability. Dörner and colleagues (Dörner and Kreuzig, 1983; Dörner, Kreuzig, Reither, and Staudel, 1983) studied individuals who were asked to play the role of city managers for the computer-simulated city of Lohhausen. A variety of problems were presented to these individuals, such as how best to raise revenue to build roads. The simulation involved more than one thousand variables. Performance was quantified in terms of a hierarchy of strategies, ranging from the simplest (trial and error) to the most complex (hypothesis testing with multiple feedback loops). No relation was found between overall cognitive ability and complexity of strategies used. A second problem was created to cross-validate these results. This problem, called the Sahara problem, required participants to determine the number of camels that could be kept alive by a small oasis. Once again, no relation was found between overall cognitive ability and complexity of strategies employed.