(r = -.34, p < .001). To examine this issue further, we created an additional summary variable
representing polarity in expressive regulation. The polarity variable was calculated as the
absolute value of the discrepancy between expressive-enhancement and expressive-suppression
ability; thus, higher polarity scores represent more extreme asymmetry in expressive regulation.
Entering the polarity variable instead of the flexibility variable in a third regression analysis did
not explain any additional variance in T2 distress and did not increase R2 beyond its value in the
first step of the analysis, F(1, 77) = 0.33, n.s. Thus, extreme ability in one form of expressive
regulation over the other does not predict change in long-term adjustment.
DISCUSSION
This study provides the first direct empirical support for the assumption that successful
adaptation is linked to the ability to flexibly enhance or suppress emotional expression. To test
this assumption, we used a within-subjects manipulation to examine expressive enhancement and
suppression as a prospective predictor of distress among New York City undergraduates
beginning college in the immediate aftermath of the September 11th terrorist attacks (T1) and
again prospectively one and a half years later (T2). Although the sample on average did not show
elevated distress, there was considerable variability in distress scores from T1 to T2. Regardless
of their level of adjustment, subjects experienced the emotion-evoking stimuli in much the same
way. However, as predicted by the expressive-flexibility hypothesis, the abilities to enhance and
to suppress the expression of emotion each independently predicted reduced T2 distress, over
and above the effects accounted for by T1 distress and the number of problems completed during
the filler task (included as a measure of individual differences in cognitive resources and
motivation). Moreover, combining expressive-enhancement and -suppression abilities into an
overall flexibility score also predicted reduced distress, whereas a polarity score, reflecting
asymmetry between expressive and suppressive abilities, was unrelated to adjustment.
At a more general level, the findings of the current study complement and extend the
corpus of studies that have examined emotion regulation using between-subjects designs. As in
the previous studies, expressive suppression did not influence the subjective experience of
emotion but attenuated the overt expression of emotion and reduced memory for emotional
stimuli. When we instructed the same subjects to enhance the expression of emotion, although
the experience of emotion was again not influenced, subjects did show greater overt displays of
emotion relative to the control condition and also exhibited a memory deficit comparable to that
observed for the suppression task, presumably because both tasks increase cognitive load.
