Is there a relationship between academic field and ability to use different types of semantic relation? Performance on two types of analogy item in the Graduate Record Examination (GRE) General Test was compared. Intensional relations are inherent in the meanings of the words and are based on shared or contrasting properties (e.g. farmer:person, breeze:gale, alive:dead, beggar:poor). Extensional relations reflect empirical relations between things in the world and are based on contiguity or causality (e.g. farmer:tractor, road:sidewalk, flu:headache).
Performance on the two kinds of analogy items was compared for a single administration of the GRE for English and History majors (verbal group, n = 2238) and electrical engineering, computer science, and mathematics majors (practical group, n = 2143). The verbal group did better on intensional, and the practical group did better on extensional analogies. The difference was not explained by a correlated gender difference by which women did better on intensional and men did better on extensional items. Thus differences in the ability to use intensional and extensional relations was related to academic training, although the direction of this relationship was not established.
Relations between ideas have been viewed as the building blocks of thought since the Greek philosophers began to ponder the nature of human knowledge. Analogy items are thought to reflect the process of recognizing or creating these relations. This assumption is one reason that analogies have been an important part of tests of verbal ability at least since they were described by Spearman (1923). If relations between concepts do play a central role in thought processes, then it is likely that an aptitude for or training in a particular type of mental ability might be reflected in facility in using a particular type of semantic relation. For example, a person with good verbal skills may be better at using some kinds of relations, while a person who is good at solving practical problems may be better at using other kinds of relations.
The present study compared the performance of two groups of Graduate Record Examination (GRE) candidates, one group majoring in fields that emphasize verbal abilities (English and history), the other majoring in fields that emphasize abilities involved in solving applied problems and manipulating sequences of events (electrical engineering, computer science, and mathematics majors). The two groups were