Abstract: Considerable research has confirmed the correlation between extraversion and positive affects (Costa et al., 1980 and 1990; Canli et al., 2001, Amin et al.,2004). In early studies, Costa and McCrae (1980) found that individuals who exhibited a high degree of extraversion—the tendency to be upbeat, optimistic, and to enjoy social contact—reported more positive emotions in everyday life than less extraverted individuals. Furthermore, this association proved effective in predicting levels of positive affects 10 years later (Costa and McCrae, 1980). In addition, highly extroverted individuals are reported slower to shift attention away from locations of positive incentive values compared with less extroverted individuals, suggesting an attentional bias of extroverts for emotionally positive stimuli (Derryberry & Reed, 1994). Consistent with these behavioral studies, the functional neuroimaging study by Amin and colleagues replicated the findings that neural activations elicited by positive stimuli increased with the degree of extroversion by an emotional dot-probe task (Amin et al., 2004). The results showed increased neural reactivity in the right fusiform gyrus with higher extroversion scores during a positive neutral match condition (the probe matched the neutral location), where extroverts made more efforts for a response due to their attentional bias for positive stimuli (Derryberry & Reed, 1994; Amin et al., 2004).
Therefore, it seems that extroverts are more sensitive to emotionally positive stimuli than less extroverted individual: they react more intensely to positive stimuli (Canli et al., 2001), and their attention is easy to be diverted by distracting stimuli laden with emotional positivity (Derryberry & Reed, 1994; Amin et al., 2004). Thus, the present study hypothesizes that extraverts direct more attention to positive stimuli, and the intensity of their attention increase with the valence strength of positive stimuli.
As paid volunteers, 14 extroverted (19–25 years; M=21.8 years, 7 males) and 14 non-extroverted (18-23 years; M=20.9 years, 7 males) students from Southwest University participated in the experiment. The subjects were selected from a large pool of 350 college students who filled in the extroversion subscale of the NEO Five-Factor Inventory (NEO-FFI). The present study adopted a modified oddball paradigm which consisted of 6 blocks of 100 trials, and each block included 70 standard and three conditions of 10 deviants. All deviants were pictures taken from the Chinese Affective Picture System (CAPS). A natural scene of cup served as the frequent standard picture and 30 pictures grouped as either highly positive (HP), moderately positive (MP), or Neutral served as the deviants. Three groups of deviant pictures were selected in such a way that they differed significantly in valence from one another, but were similar in arousal. Each subject was instructed to press the “F” key on the keyboard as accurately and quickly as possible if the standard picture appeared, and to press the “J” key if the deviant picture appeared. The EEG for correct response during each valence condition was overlapped and averaged separately.
At about 170ms after stimulus onset, a prominent P2 component was elicited in all three valence conditions, and P2 amplitudes were largest at central and central-frontal sites, irrespective of valence condition. We found a significant valence and extroversion interaction effect on P2 amplitudes. The results showed that in extroverts, highly positive stimuli elicited larger P2 amplitudes than moderately positive stimuli which, in turn, elicited larger amplitudes than the neutral stimuli. In contrast, ERPs were similar across the three valence conditions in non-extraverts.
The result indicates that the attentional bias of extraverts for positive stimuli increases with the valence intensity of positive stimuli

Key words: extraverts, attentional bias, positive stimuli, ERP, valence intensity