the cognitive segment. Some data explicitly link extraversion–introversion to
intellectual performance. Several studies show that extraverts perform better
than introverts on intelligence tests under conditions of high stress or arousal,
although the extraversion–arousal interaction reverses in the evening (Re-
velle, 1993). These results are paralleled by similar interactive effects on infor-
mation-processing tasks requiring semantic processing (Matthews & Harley,
1993). There are also stylistic differences between extraverts and introverts,^1
in that extraverts are poorer at reflective problem solving, because they exit
from the problem prematurely (Matthews, 1997).
A second observation is that the magnitude of correlations between
extraversion and behavioral criteria is typically quite small (e.g., 0.2–0.3).
The paradox of personality is that constructs such as those of the Big Five
(Goldberg, 1993) emerge very strongly in psychometric studies, and yet the
role of personality is often elusive when we look at physiological and cogni-
tive functioning in controlled experiments (Matthews & Gilliland, 1999). We
note briefly that we reject the view that extraversion is essentially dis-
positional positive affect (Watson, 2000). Studies of mood in controlled envi-
ronments show that correlations between extraversion and affective states
are typically around 0.1–0.4 (Matthews & Gilliland, 1999). Extraversion may
indeed relate to general life satisfaction, but this itself is a complex construct
with multiple facets. The picture that emerges from empirical studies (e.g.,
Matthews, 1997) is that there is no single master process that determines level
of extraversion–introversion, irrespective of whether we look at psycho-
physiological constructs, information processing or high-level goals, and self-
knowledge. Instead, extraversion is distributed across multiple processes at
different levels of abstraction.
With regard to the trilogy, the higher level correlates of extraversion in-
clude dispositional happiness (emotion), self-efficacy beliefs (cognition), and
social interests (motivation). Lower-level correlates include more positive
moods (emotion), sensitivity of brain mechanisms for reward (motivation),
and biases in attentional and memory processes (cognition). Even at the level
of molecular genetics, the emerging evidence suggests that extraversion will
relate in small ways to many genes (e.g., Plomin & Caspi, 1998).
The multifarious correlates of extraversion present a unique challenge for
theory. The simple approach, of finding one key physiological or psychologi-
cal process that will explain all the data, seems unlikely to succeed. However,
an adaptive perspective may be more productive. Many of the correlates of
extraversion appear to point in the same direction—toward an adaptation
for demanding social environments (Matthews, 1997; Matthews & Dorn,
148 MATTHEWS AND ZEIDNER
1 1 It is convenient to contrast behavioral differences between ‘extraverts’ and ‘introverts’, but
note that extraversion–introversion is a continuous variable, not a typology.