Although Jones and Day's (1997) findings suggest that there are different processes associated with solving novel social problems than those used to solve familiar social problems or novel academic problems, there are some limitations to their study. First, the sample (i.e., high school students) may represent individuals who are relative novices when it comes to social problem solving, such that their level of knowledge may reflect abstract concepts that are similar to academic-type problems. Individuals who have more expertise in social problems may have knowledge that is more distinct from academic problem-solving skill. Second, the method of measuring each of these factors may have contributed to the findings. Both Crystallized Social Knowledge and Academic Problem Solving involved items with one correct answer, whereas the measures of Social-Cognitive Flexibility asked respondents to provide their own interpretation, rate the importance of different social goals, and to identify the most effective solution to achieve the social goal. The similarity in the measurement format for the former two measures may have created an artificially higher validity estimate among them.

The limitations identified by Jones and Day (1997) are some of the concerns that Kihlstrom and Cantor (in press) raise about relying on psychometric approaches to study social cognition. Cantor, Kihlstrom and colleagues (Cantor and Harlow, 1994; Cantor and Kihlstrom, 1987; Kihlstrom and Cantor, in press) take a social-cognition view of personality. These researchers do agree that social behavior is intelligent because it is mediated by cognitive processes such as perception, memory, reasoning, and problem solving. They argue that psychometric approaches to understanding social cognition inappropriately focus on how much social cognition a person has rather than what social cognition the person possesses. Individual differences in social behavior can be attributed to differences in knowledge and strategies needed to accomplish social tasks.

Cantor and Harlow (1994) proposed that intelligent behavior involves attunement to the consequences of one's actions, the implications of those consequences for other goals, and the goal-fulfilling potentials of different situations. Attunement allows for flexibility in terms of what tasks to pursue, where and when opportunities are present to work on various tasks, and how to pursue the tasks. Therefore, attunement and flexibility are critical aspects of personality and cognition, allowing individuals successfully to pursue goal and solve problems. Cantor and Harlow argued that, due to the varied tasks and settings in which individuals behave, it is difficult to obtain a general, stable measure of social cognition.

Rather than developing instruments to assess individual differences in social cognition, Cantor and her colleagues (see Kihlstrom and Cantor, in press) have chosen to study the cognitive processes that support intelligent social behavior. They focus on life tasks as their unit of analysis for studying social cognition. Life tasks are identified by the individual as meaningful and serve to organize one's daily activities. They allow researchers to observe the skill of people to solve problems of a social nature and the knowledge they have of how to solve them. Life tasks include things like making friends, finding a spouse, establishing a career, and getting good grades. Cantor and her colleagues have chosen to focus on periods of transition (e.g., from high school to college) to observe individual differences in life tasks (see Kihlstrom and Cantor, in press). They have found that people formulate action plans, monitor their progress, and assess the outcomes of their actions. They draw on their autobiographical memory to evaluate various causes of those outcomes and alternative courses of action. When their pursuit of a life task is obstructed, they are able to alter plans or choose new plan of action. As we discuss in a later part of this section, the processes identified by Cantor and her colleagues are consistent with the metacomponents identified by Sternberg (1985a) as underlying successful cognition.