This investigation yielded two main results. First, participants presented a general tendency to perceive neutral faces as negative and an overall emotional bias towards sadness. Higher levels of resiliency were associated with judging neutral faces less negatively, and presenting a greater general bias towards happiness when compared to low resilient participants. Second, trait anxiety, state depression, extraversion and neuroticism were unable to account for the relationship between resilience and the (less negative) judgment of neutral faces. Taken together, a positive bias in high resilient individuals during emotion judgment and an attenuated perception of perceiving neutral expressions as negative, may provide insight into how resilient individuals engage cognitive and affective processes to decode emotional aspects of facial expressions. This altered engagement may contribute to their efficient adaptation in difficult interpersonal situations.
Our results indicate a general bias towards viewing neutral faces as sad, which is consistent with previous investigations that have reported that, when a neutral face is mislabeled, it is often considered sad (Gur et al., 2002
;Rojahn et al., 1997
). Some investigators have suggested that the context in which neutral faces are presented can influence judgment so that neutral faces presented among happy faces are considered sad (Russell, 1991
). That is, it is not only the currently processed stimulus but also its preceding context can influence this emotion judgment (Kuleshov effect; Prince et al., 1992
). However, this would not be a plausible explanation for the current results since our stimuli were presented in a pseudorandom order to assure equal number of neutral faces after a happy, sad, and fearful expression.
In the current study, this overall sadness bias, particularly during neutral face evaluation, was further related to the degree of self-reported resilience, resulting in a greater tendency toward happiness perception and reduced sadness bias in highly resilient individuals during H-F and S-H pairings. Taking into account the entire sample, we also observed a positive relationship between resilience and happiness perception while evaluating neutral faces. These results are consistent with observations that resilient individuals are able to generate positive emotions to help them cope with stressful situations (Tugade, Fredrickson, and Barrett, 2004
). According to Fredrickson's broaden-and-build theory, positive emotions facilitate enduring personal resources and broaden one's momentary thought of action repertoire (Fredrickson, 2004
). That is, positive emotions broaden one's awareness and encourage novel, varied, and exploratory thoughts and actions which, in turn, build skills and resources. For example, experiencing a pleasant interaction with a person you asked for directions turns, over time, into a supportive friendship. Furthermore, positive emotions help resilient individuals to achieve effective coping (Werner et al., 1992
) serving to moderate stress reactivity and mediate stress recovery (Ong et al., 2006
). We suggest as an explanation for our findings that individuals high on self-reported resilience may be more likely to process information that is congruent with a positive view of the world, and that this capacity helps maintain their homeostasis. This positive bias during emotion perception may provide the rose-colored glasses that resilient individuals use to interpret the world and achieve effective ways to bounce back from adversity (Bonanno, 2004
) and maintain wellness.
We explored whether the relationship between resilience and bias as well as the perception of neutral faces were better accounted for by trait anxiety, extraversion, neuroticism, or by current levels of depression. Results revealed that, on the one hand, resilience was the best predictor of neutral face rating and, on the other hand, extraversion was the best predictor for bias. Thus, despite a strong negative correlation between resilience and anxiety (see also Campbell-Sills, Cohan, and Stein, 2006
), depression and neuroticism as well as a positive relationship with extraversion, our results suggest that: 1) extraversion may play an important role for an overall bias towards positive emotion and 2) resilience may be a more specific component in the face of ambiguity (i.e., neutral faces). Thus, we can conclude that the construct of resilience may offer a perspective on physical and psychological response to stress that is not the mere inverse of psychopathology.
There are several limitations to the current study. We are aware of cross-cultural differences in the agreement of facially depicted emotions both in the receiver (Hart et al., 2000
) and sender (Hess et al., 2000
). Future studies should consider paradigms including multi-ethnic participants/actors as well as incorporate other positive and negative stimuli to rule out whether this effect is specific to facial stimuli. Taking into account the limitations of a self-report instrument, our measure of resilience may be biased in the direction of social desirability. Experimental manipulation of stress induction followed by task performance (e.g., facial perception of emotion) should be developed as an in vivo
validation of an objective measure of “online” resilience. Due to task limitations, the current paradigm included an unequal number of positive (one, happy) and negative (two, sad and fearful) emotions that may have partially contributed to a general negative bias response. Nevertheless, this limitation should not account for the effect of resilience on emotional bias. Finally, our study did not include a “neutral emotion” response alternative. This is an inevitable consequence of studying emotional bias in perception. However, a previous study in which neutral was a response option (Rojahn and Warren, 1997
) reported that when participants mislabeled the emotion, there was a bias towards negative affect (i.e., sad or angry).
The fact that even non-treatment seeking volunteers do not consider “neutral” faces equally often as happy or sad has important repercussions in the study of emotions using subtraction techniques. That is, in functional neuroimaging studies, when comparing positive or negative to neutral facial emotions, the non-neutrality of neutral faces may be eliminating the effect of sad, and polarizing that of positive expressions, respectively. Moreover, amygdala activation has been associated with novel (Schwartz et al., 2003
) and unfamiliar (Hart, Whalen, Shin, McInerney, Fischer, and Rauch, 2000
) faces claiming the role of this limbic structure in salient stimulus processing (Wright et al., 2006
). Although this may be a plausible explanation, it should be clarified whether these faces were considered neutral and not negatively valenced. Future experiments may implement a baseline or comparator condition that does not involve faces (e.g., shapes).
In summary, our study suggests that resilience may be associated with positive emotional perception as reflected by an attenuated bias towards negative affect. That is, resilience – and not other traits such as neuroticism or anxiety or states such as depression – may provide the rose-colored glasses used when individuals are forced to make an affective assessment in the context of an uncertain emotional situation. Taken together, these data have practical implications for studies using neutral faces as a comparator to several emotions and, moreover, suggest the importance of resilience and its influence on the way we perceive our interpersonal world.