8 Personality and Social Psychology Bullet<strong>in</strong> XX(X)and sadness. 4 Th<strong>at</strong> is, the gre<strong>at</strong>er <strong>in</strong>hibition and lower approachmotiv<strong>at</strong>ion associ<strong>at</strong>ed with sadness, r<strong>at</strong>her than the contrast<strong>in</strong>gagency appraisals, may have led participants to subsequentlyexperience less anger because <strong>of</strong> the contrast<strong>in</strong>gmotiv<strong>at</strong>ional systems associ<strong>at</strong>ed with anger. To test this possibility,we assessed the moder<strong>at</strong><strong>in</strong>g role <strong>of</strong> BAS (BIS) on theemotional blunt<strong>in</strong>g <strong>of</strong> anger (sadness).A third objective <strong>of</strong> Study 2 was to determ<strong>in</strong>e the bidirectionality<strong>of</strong> emotional blunt<strong>in</strong>g. Th<strong>at</strong> is, would anger blunt theelicit<strong>at</strong>ion <strong>of</strong> sadness just as sadness blunts the elicit<strong>at</strong>ion <strong>of</strong>anger? Our framework suggests th<strong>at</strong> the underly<strong>in</strong>g agencyappraisals can be <strong>in</strong>fluential irrespective <strong>of</strong> their direction.Thus, we aimed to demonstr<strong>at</strong>e th<strong>at</strong> emotional blunt<strong>in</strong>geffects are bidirectional for both subsequent emotions andcognitions. 5F<strong>in</strong>ally, one may wonder wh<strong>at</strong> occurs when <strong>in</strong>dividualstransition from one emotion-elicit<strong>in</strong>g event to another eventth<strong>at</strong> elicits the same emotion. Because the subsequent emotionaltre<strong>at</strong>ment <strong>in</strong>evitably re<strong>in</strong>forces the same cognitiveappraisals as the prior emotional experience, emotional augment<strong>at</strong>ion(Ames & Johar, 2009), the opposite <strong>of</strong> emotionalblunt<strong>in</strong>g, could occur. To clarify this question, the presentstudy employed a fully-crossed design with not only angerto-sadnessand sadness-to-anger transitions, but also angerto-angerand sadness-to-sadness transitions.MethodParticipants and procedure. One-hundred and n<strong>in</strong>ety-threeundergradu<strong>at</strong>e students (91 females, 102 males) completedthe study for course credit. We followed a “two-study”cover story <strong>in</strong> which participants were randomly assigned toone <strong>of</strong> six conditions <strong>in</strong> a 3 (emotion <strong>in</strong>duction: sadness,anger, neutral) 2 (subsequent emotion tre<strong>at</strong>ment: sadness vs.anger) fully-crossed between-subjects design with BIS/BASscales measured as cont<strong>in</strong>uous variables.Basel<strong>in</strong>e emotion and BIS/BAS scales were completed <strong>at</strong>the beg<strong>in</strong>n<strong>in</strong>g <strong>of</strong> a 1-hr research session. <strong>The</strong> scales conta<strong>in</strong>ed20 items, and response options on the BIS/BAS scalesranged from 1 (very false for me) to 7 (very true for me). Onedimension (7 items, α .82) reflects BIS sensitivity, or thre<strong>at</strong>responsiveness, whereas the other three dimensions (theaverage correl<strong>at</strong>ion <strong>in</strong> this sample was .42) reflect aspects <strong>of</strong>BAS sensitivity, or <strong>in</strong>centive responsiveness: Fun Seek<strong>in</strong>g(4 items, α .88), Drive (4 items, α .88), and RewardResponsiveness (5 items, α .85). <strong>The</strong> BIS dimension was<strong>of</strong> primary <strong>in</strong>terest for emotional blunt<strong>in</strong>g <strong>of</strong> sadness, whereasthe BAS Drive dimension r<strong>at</strong>her than Fun Seek<strong>in</strong>g or RewardResponsiveness was the focus for emotional blunt<strong>in</strong>g <strong>of</strong>anger (Carver, 2004; Harmon-Jones, 2003; Wilkowski &Meier, 2010).<strong>The</strong> procedure is detailed <strong>in</strong> Figure 2. Approxim<strong>at</strong>ely30 m<strong>in</strong> after the start <strong>of</strong> the research session, participantscompleted the two-question essay writ<strong>in</strong>g task (Lerner &Table 2. Agency Appraisal Items (Studies 1 and 3)Someone else had the ability to <strong>in</strong>fluence wh<strong>at</strong> was happen<strong>in</strong>g. (H)Another <strong>in</strong>dividual was to blame for wh<strong>at</strong> was happen<strong>in</strong>g. (H)<strong>The</strong> events were caused by human control. (H)No <strong>in</strong>dividual was to blame for wh<strong>at</strong> was happen<strong>in</strong>g. (S)<strong>The</strong> events were caused by situ<strong>at</strong>ional factors. (S)<strong>The</strong> events were beyond any human’s control. (S)<strong>The</strong> agency appraisal items were adapted from Smith and Ellsworth(1985). Responses ranged from 1 (strongly disagree) to 7 (strongly agree).<strong>The</strong> three human agency items ( Study 2 .83, Study 3 .95) and twositu<strong>at</strong>ional agency items were averaged to cre<strong>at</strong>e a composite score( Study 2 .74, Study 3 .94). We used only two <strong>of</strong> the three situ<strong>at</strong>ionalagency items because factor analysis <strong>in</strong>dic<strong>at</strong>ed the second situ<strong>at</strong>ionalagency item did not factor well; reliability <strong>of</strong> the <strong>in</strong>dex <strong>in</strong>creased withexclusion <strong>of</strong> the second situ<strong>at</strong>ional agency item ( Study 2 .74 from. 63, Study 3 .94 from .85). H human; S situ<strong>at</strong>ional.Keltner, 2001), which was followed by the emotion selfreportand appraisal measures (see Table 2). <strong>The</strong>n, to elicit asubsequent emotion experience to assess emotional blunt<strong>in</strong>g,participants next read a scenario from Keltner, Ellsworth, et al.(1993) and responded to the emotion list and Life EventsQuestionnaire. <strong>The</strong> three anger measures (angry, irrit<strong>at</strong>ed,and mad; α .93) and the three sadness measures (sad,depressed, and downhearted; α .89) were <strong>of</strong> primary <strong>in</strong>terest.Optimistic risk estim<strong>at</strong>es were assessed with the sameLife Events Questionnaire from Study 1 (eigenvalue 1.91;α .69).Manipul<strong>at</strong>ion check. Self-reported emotions after the emotion<strong>in</strong>duction revealed th<strong>at</strong> anger condition participantsreported significantly more anger than did sadness conditionparticipants (Ms 5.87 vs. 3.03), t 6.61, p .01, orneutral condition participants (M 2.28), t 8.93, p .01.In contrast, sadness condition participants reported significantlymore sadness than did anger condition participants(Ms 6.08 vs. 3.11), t 7.55, p .01, or neutral conditionparticipants (M 1.61), t 12.40, p .01. Thus, our emotion<strong>in</strong>duction worked as anticip<strong>at</strong>ed.Results and DiscussionSubsequent Anger Tre<strong>at</strong>ment ResultsInferential analysis: Blunt<strong>in</strong>g anger. We proposed th<strong>at</strong> <strong>in</strong>dividualsexperienc<strong>in</strong>g sadness (vs. neutral or angry st<strong>at</strong>es)would subsequently experience less anger. However, we alsotheorized th<strong>at</strong> the motiv<strong>at</strong>ional approach system (accessedvia BAS Drive) could moder<strong>at</strong>e the blunt<strong>in</strong>g effect <strong>of</strong> sadnesson anger. To test the hypothesized moder<strong>at</strong><strong>in</strong>g role <strong>of</strong>BAS Drive, we conducted a regression for those with angeras the subsequent emotion tre<strong>at</strong>ment. Emotion <strong>in</strong>duction,BAS Drive, and their <strong>in</strong>teraction were the <strong>in</strong>dependent variables;self-reported composite anger score was the dependentvariable; and basel<strong>in</strong>e anger and sadness were <strong>in</strong>cludedas covari<strong>at</strong>es. 6 All variables were mean-centered before
W<strong>in</strong>terich et al. 9Table 3. Regression Analyses for Subsequent Anger (Sadness) Tre<strong>at</strong>ments: Self-Reported Anger (Sadness) and Optimistic RiskEstim<strong>at</strong>es (Study 2)Subsequent anger tre<strong>at</strong>ment resultsAngerOptimistic risk estim<strong>at</strong>esVariable B SE B B SE B Sadness emotion .05 .14 .05 .48 .25 .23*Neutral emotion .17 .13 .15 .02 .24 .01BAS Drive .17 .10 .29* .02 .17 .02Sadness BAS .40 .15 .34† .47 .27 .23*Neutral BAS .34 .13 .39† .42 .23 .27*Basel<strong>in</strong>e sadness .06 .04 .20* .02 .07 .04Basel<strong>in</strong>e anger .02 .05 .05 .07 .08 .11Medi<strong>at</strong>ion analysisSadness BAS .22 .17 .18Neutral BAS .15 .15 .18Human agency appraisals .05 .07 .08Human Agency <strong>Appraisals</strong> BAS .15 .07 .28†Subsequent sadness tre<strong>at</strong>ment resultsSadnessOptimistic risk estim<strong>at</strong>esVariable B SE B B SE B Anger emotion .66 .46 .18 .32 .23 .17Neutral emotion .17 .42 .05 .09 .21 .05BIS .27 .35 .15 .31 .18 .34*Anger BIS .97 .44 .35† .56 .22 .40†Neutral BIS .97 .46 .32† .51 .24 .32†Basel<strong>in</strong>e sadness .09 .13 .07 .10 .07 .17Basel<strong>in</strong>e anger .14 .12 .13 .10 .06 .17Medi<strong>at</strong>ion analysisAnger BIS .27 .55 .10Neutral BIS .73 .47 .24*Situ<strong>at</strong>ional agency appraisals .27 .11 .29†Situ<strong>at</strong>ional Agency <strong>Appraisals</strong> BAS .26 .13 .57†BIS behavioral <strong>in</strong>hibition system; BAS behavioral approach system.†p .10. *p .05. [AQ: 6]the analysis. We cre<strong>at</strong>ed two dummy variables for emotion<strong>in</strong>duction because it is a three-level c<strong>at</strong>egorical variable: sad,anger, and neutral (Aiken & West, 1991). Anger was the basecomparison condition with a dummy variable for sadnessand neutral conditions. <strong>The</strong> sadness and neutral dummy variableswere each <strong>in</strong>teracted with BAS Drive.Results revealed th<strong>at</strong>, consistent with our theoriz<strong>in</strong>g, the<strong>in</strong>teraction <strong>of</strong> the sadness condition and BAS Drive was significant,β .34 (.15), t 2.59, p .05, as was the <strong>in</strong>teractionbetween the neutral condition and BAS Drive, β .39 (.13),t 2.60, p .05. See Table 3 for complete regression analysisresults. To exam<strong>in</strong>e whether the p<strong>at</strong>tern <strong>of</strong> these <strong>in</strong>teractionsis consistent with our predictions, we conducted simpleslope analysis <strong>at</strong> 1 SD below the mean <strong>of</strong> BAS Drive. <strong>The</strong>key question <strong>of</strong> <strong>in</strong>terest is whether anger is blunted whenparticipants with low BAS Drive are experienc<strong>in</strong>g sadnessrel<strong>at</strong>ive to when participants with low BAS Drive are experienc<strong>in</strong>ganger. As hypothesized, sadness condition participantsreported less anger after the subsequent anger tre<strong>at</strong>mentthan did anger condition participants, β –.43 (.23), t 2.24,p .05, d .47, <strong>in</strong>dic<strong>at</strong><strong>in</strong>g th<strong>at</strong> sadness blunts the subsequentexperience <strong>of</strong> anger for <strong>in</strong>dividuals with low BAS. Susceptibilityto the anger tre<strong>at</strong>ment did not differ as a function <strong>of</strong>whether participants had been exposed to a neutral <strong>in</strong>ductionor to an anger <strong>in</strong>duction, β –.19 (.19), t –1.14, p .20,d .24, which suggests th<strong>at</strong> it is the specific experience <strong>of</strong>sadness r<strong>at</strong>her than the experience <strong>of</strong> any emotion (i.e., anger)th<strong>at</strong> blunts anger. We f<strong>in</strong>d no evidence for emotional blunt<strong>in</strong>g<strong>at</strong> high levels (1 SD above the mean) <strong>of</strong> BAS Drive, asexpected. Similar to Study 1, these f<strong>in</strong>d<strong>in</strong>gs support our theoryth<strong>at</strong> participants experienc<strong>in</strong>g sadness subsequentlyexperience less anger rel<strong>at</strong>ive to those first experienc<strong>in</strong>g
Change language