Throughout its history, the study of social cognition has periodically fallen out of favor with researchers. This lack of interest can be attributed to failed attempts to distinguish measures of social from measures of abstract cognition. The difficulty in distinguishing social from academic or abstract cognition can be explained by efforts that focus primarily on cognitive aspects of social cognition and methods that rely heavily on verbal assessment. Researchers as early as Thorndike (1920) acknowledged the multidimensional nature of social cognition. Until recently, however, the approaches to studying social cognition have emphasized cognitive aspects, such as social perception (e.g., Chapin, 1942) and moral reasoning (e.g., Keating, 1978). In order to assess these cognitive dimensions, researchers relied, to a large extent, on verbal measures. Measures of behavioral aspects of social cognition also have relied somewhat on verbal forms of assessment (e.g., self-report). As becomes clear from a brief review of the literature, research efforts that consider behavioral and nonverbal measures of social cognition have had greater success in establishing discriminant validity from measures of abstract cognition than have the more cognitive, verbal measures of social cognition.

2.1.1 Cognitive-verbal measures of social cognition

Many approaches to understanding social cognition follow the tradition of cognition testing by developing instruments to assess individual differences in social cognition. One of the first and better known tests of social cognition was the George Washington Social Cognition Test (GWSIT; Moss, Hunt, Omwake, and Woodward, 1949). This test consists of a number of subtests that assess judgment in social situations, recognition of the mental states behind messages, memory for names and faces, observation of human behavior, and sense of humor. Early research with the GWSIT suggested that it could not be distinguished easily from abstract cognition (e.g., Thorndike and Stein, 1937).

A set of social-cognition tests emerged within the context of Guilford's (1967) Structure of Intellect Model of Cognition. Within Guilford's framework, social cognition is viewed as comprising those skills within the domain of behavioral operations. O'Sullivan, Guilford, and deMille (1965) developed tests to measure behavioral cognition, which they defined as the skill to judge people. More specifically, the tests measured the skill to decode social cues, including facial expressions, vocal inflections, posture, and gestures. In a study with 306 high-school students, O'Sullivan et al. (1965) found evidence that their factors of social cognition were distinct from measures of abstract cognitive skill. Later research, however, found contradictory results (e.g., Riggio, Messamer, and Throckmorton, 1991).

Riggio et al. (1991) administered several measures of social cognition and several measures of academic cognition to undergraduate students. Academic cognition was measured using the Shipley-Hartford Institute of Living Scale (Shipley, 1940), which measures verbal and abstract reasoning, and the vocabulary subscale of the WAIS-R (Wechsler, 1981). Measures of social cognition included four tests of the Factor Tests of Social Cognition (O'Sullivan and Guilford, 1976); Riggio's (1986, 1989) Social Skills Inventory (SSI), which assess six social communication skills (emotional expressivity, emotional sensitivity, emotional control, social expressivity, social sensitivity, and social control); and a social etiquette/tacit knowledge test that measured knowledge of appropriate behaviors in social situations. Riggio et al. found comparable intercorrelations within measures of both academic and social cognition as they did between measures of academic and social cognition. An exploratory factor analysis suggested two factors, one that included the Shipley-Hartford Abstract Reasoning scale and the Guilford measures and was labeled "abstract reasoning cognition," and the second that included the Shipley-Hartford Verbal scale and the SSI, which was labeled "verbal cognition." These findings suggested that academic and social cognition are overlapping domains. At the same time, these researchers found little evidence of convergent validity among the measures of social cognition, likely reflecting the complexity of the construct and the various ways it has been operationalized in the literature.