Anda di halaman 1dari 22

ARTICLE IN PRESS

ARTICLE IN PRESS Journal of Research in Personality xxx (2006) xxx–xxx www.elsevier.com/locate/jrp 1 2 3 4r providing the aggre ssion paradigm information. We thank Ryan Adams, Gary Lautenschlager, William Graziano, J ames E r i c k son, James Long, Anita Miller, and Wendi Gardner for their comments and advice. We thank the UTA Personality and Social Behavior Lab, especially Nichole Bryant, Katie Gleason, Sarah Chan, Melissa Cook, Jamie Jones, Owen Temple, and Larry Carter. We also thank Andres Campbell for his countless hours as a practice participant. Corresponding author. Fax: +1 817 272 2364. E-mail address: Lcampbell@uta.edu (L.A. Jensen-Campbell). 0092-6566/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.jrp.2006.05.001 " id="pdf-obj-0-6" src="pdf-obj-0-6.jpg">
ARTICLE IN PRESS Journal of Research in Personality xxx (2006) xxx–xxx www.elsevier.com/locate/jrp 1 2 3 4r providing the aggre ssion paradigm information. We thank Ryan Adams, Gary Lautenschlager, William Graziano, J ames E r i c k son, James Long, Anita Miller, and Wendi Gardner for their comments and advice. We thank the UTA Personality and Social Behavior Lab, especially Nichole Bryant, Katie Gleason, Sarah Chan, Melissa Cook, Jamie Jones, Owen Temple, and Larry Carter. We also thank Andres Campbell for his countless hours as a practice participant. Corresponding author. Fax: +1 817 272 2364. E-mail address: Lcampbell@uta.edu (L.A. Jensen-Campbell). 0092-6566/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.jrp.2006.05.001 " id="pdf-obj-0-8" src="pdf-obj-0-8.jpg">

Journal of Research in Personality xxx (2006) xxx–xxx

ARTICLE IN PRESS Journal of Research in Personality xxx (2006) xxx–xxx www.elsevier.com/locate/jrp 1 2 3 4r providing the aggre ssion paradigm information. We thank Ryan Adams, Gary Lautenschlager, William Graziano, J ames E r i c k son, James Long, Anita Miller, and Wendi Gardner for their comments and advice. We thank the UTA Personality and Social Behavior Lab, especially Nichole Bryant, Katie Gleason, Sarah Chan, Melissa Cook, Jamie Jones, Owen Temple, and Larry Carter. We also thank Andres Campbell for his countless hours as a practice participant. Corresponding author. Fax: +1 817 272 2364. E-mail address: Lcampbell@uta.edu (L.A. Jensen-Campbell). 0092-6566/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.jrp.2006.05.001 " id="pdf-obj-0-12" src="pdf-obj-0-12.jpg">

www.elsevier.com/locate/jrp

1

2

3

4

5

6

7

Do Big Five personality traits associated with

self-control inXuence the regulation of anger

and aggression?

Lauri A. Jensen-Campbell ¤ , Jennifer M. Knack,

Amy M. Waldrip, Shaun D. Campbell

Department of Psychology, University of Texas at Arlington, Box 19528, Arlington, TX 76019-0528, USA

  • 8 Abstract

  • 9 This study examined whether Big Five personality traits associated with the ability to exhibit self-

    • 10 control would moderate the anger–aggression link. A total of 126 participants (63 women) completed

    • 11 personality measures. In a separate experimental session, participants wrote an essay and then received

    • 12 either positive or negative feedback from a Wctitious participant. Participants were given the opportu-

    • 13 nity to aggress against the supposed other person. Baseline and post-experimental emotions were

    • 14 assessed. EEG was recorded to measure activity in midfrontal, lateral-frontal, and parietal areas.

    • 15 Results replicated previous Wndings that anger is associated with left relative to right prefrontal asym-

    • 16 metry and aggression. Conscientiousness was negatively associated with anger and relative left prefron-

    • 17 tal asymmetry. Conscientiousness also moderated the link between anger and aggression. Agreeableness

    • 18 was positively associated with anger, but only when levels of conscientiousness were low.

    • 19 © 2006 Elsevier Inc. All rights reserved.

    • 20 Keywords: Big Five; Agreeableness; Conscientiousness; Anger regulation; Prefrontal asymmetry; Aggression

We thank Eddie Harmon-Jones for providing the aggression paradigm information. We thank Ryan Adams, Gary Lautenschlager, William Graziano, James Erickson, James Long, Anita Miller, and Wendi Gardner for their comments and advice. We thank the UTA Personality and Social Behavior Lab, especially Nichole Bryant, Katie Gleason, Sarah Chan, Melissa Cook, Jamie Jones, Owen Temple, and Larry Carter. We also thank Andres Campbell for his countless hours as a practice participant. * Corresponding author. Fax: +1 817 272 2364. E-mail address: Lcampbell@uta.edu (L.A. Jensen-Campbell).

0092-6566/$ - see front matter © 2006 Elsevier Inc. All rights reserved.

doi:10.1016/j.jrp.2006.05.001

ARTICLE IN PRESS

  • 2 L.A. Jensen-Campbell et al. / Journal of Research in Personality xxx (2006) xxx–xxx

  • 21 1. Introduction

  • 22 In every society, self-control is important for getting along with others. It has even been

  • 23 suggested that the overarching purpose of self-control in humans is inherently social in

  • 24 nature (Barkley, 2001). A person who cannot control his or her thoughts, feelings, or

  • 25 behaviors is more likely to lash out in anger when frustrated, handle conXicts less construc-

  • 26 tively, and engage in antisocial behavior (Baumeister & Vohs, 2004). Even nonaggressive

  • 27 children who have attentional problems often have diYculty giving situationally appropri-

  • 28 ate responses and have diYculty modulating their behavior to meet the social demands of

  • 29 the situation (Landau & Milich, 1988). Conversely, individuals who are able to control

  • 30 their behaviors should receive numerous beneWts from their social relations within the

  • 31 group such as belongingness, peer acceptance, higher quality relationships, and even pro-

  • 32 tection from victimization (Jensen-Campbell & Malcolm, 2004; Vohs & Ciarocco, 2004).

  • 33 Research has indeed found associations between attentional processes, emotion-related

  • 34 regulation, and social behavior. For example, preschoolers’ attentional control was associ-

  • 35 ated with constructive anger reactions (Eisenberg, Fabes, Nyman, Bernzweig, & Pinuelas,

  • 36 1994). Additionally, Eisenberg et al. (1995) found that high attentional control in elemen-

  • 37 tary school boys predicted socially appropriate behavior. Teacher and parent reports of

  • 38 behavioral regulation were also associated with lower levels of aggressive and disruptive

  • 39 behavior. Similarly, Belsky, Friedman, and Hsieh (2001) found that early attentional con-

  • 40 trol moderates the association between negative emotionality and social competence.

  • 41 The purpose of this study was to examine whether Big Five personality dimensions

  • 42 associated with self-control, namely agreeableness and conscientiousness, inXuence emo-

  • 43 tional and behavioral responses when individuals were confronted with a frustrating inter-

  • 44 personal situation. If frustration is a major instigator of aggression (Berkowitz, 1968), an

  • 45 important mechanism that inhibits aggressive behavior may be the ability to exhibit self-

  • 46 control in frustrating situations. Many studies have found that anger is often directly asso-

  • 47 ciated with aggression when other defensive motivations are absent (e.g., Berkowitz, 1993;

  • 48 Richardson et al., 1998). For example, Harmon-Jones and Sigelman (2001) have found that

  • 49 relative left prefrontal cortical activity is associated with state-induced anger, which in turn

  • 50 predicted oVensive aggression.

  • 51 Previous research has focused primarily on “main eVect” models of anger and aggres-

  • 52 sion (Anderson & Bushman, 2002; Anderson & Carnagey, 2004). Interactive models, on

  • 53 the other hand, allow for the exploration of moderator or “buVer” variables that inXuence

  • 54 these associations. We were speciWcally interested in whether conscientiousness and agree-

  • 55 ableness moderated the contributions of anger to aggression. Individuals who have the

  • 56 ability to eVortfully shift and focus their attention may be less likely to aggress when

  • 57 angered than individuals who have less self-control. In other words, the behavioral and

  • 58 cognitive control capacities associated with conscientiousness and agreeableness may aid

  • 59 in weakening the anger–aggression link. Agreeable and conscientious individuals may be

  • 60 better able to suppress the dominant or prepotent response to be angry and oVensively

  • 61 aggressive when involved in a frustrating situation.

  • 62 1.1. Self-control and personality

  • 63 Rothbart and Bates (1998) and Rothbart and Posner (1985) suggest that self-regulation,

  • 64 also termed eVortful control (EC), involves constitutionally based executive control

ARTICLE IN PRESS

L.A. Jensen-Campbell et al. / Journal of Research in Personality xxx (2006) xxx–xxx

3

  • 65 processes that regulate behavior reactivity (e.g., emotionality). Early appearing individual

  • 66 diVerences in the ability to sustain and shift attention as well as the ability to initiate and

  • 67 inhibit actions voluntarily have been associated with EC. EVortful control has been further

  • 68 associated with the ability to suppress a dominant or prepotent behavior (Kochanska,

  • 69 Murray, & Harlan, 2000) or even another opposing dominant response that would better

  • 70 maintain subjective well being (Larsen & Prizmic, 2004). Overall, self-control should allow

  • 71 individuals to resist the immediate inXuences of the situation; individuals higher on self-

  • 72 control should be able to control their emotional aVect (e.g., anger), avoid lashing out

  • 73 in situations that are frustrating, or even approach frightful situations when necessary.

  • 74 A developmental connection between early appearing self-control processes and later

  • 75 personality processes has been documented in numerous studies. Rothbart, Chew, and

  • 76 Gartstein (2001) have found that conscientiousness is highly correlated with the adult tem-

  • 77 peramental factor of eVortful attention. Ahadi and Rothbart (1994) note that conscien-

  • 78 tiousness is composed of numerous characteristics associated with self-regulation. For

  • 79 example, both conscientiousness and self-regulation are characterized by the ability to

  • 80 eVortfully direct one’s attention, the ability to inhibit certain behaviors to execute alternate

  • 81 behaviors, and the ability to persist in tasks. Conscientiousness has also been related to a

  • 82 broad band of performance standards, including elementary school grades (e.g., Digman &

  • 83 Inouye, 1986), middle school grades (Jensen-Campbell, 2006), performance at work and

  • 84 industrial settings (e.g., Barrick & Mount, 1991), driving accident involvement (Arthur &

  • 85 Graziano, 1996), prosocial behavior (Abe, 2005), and even success in interpersonal rela-

  • 86 tionships (Jensen-Campbell & Malcolm, 2006). Furthermore, people high in conscientious-

  • 87 ness are able to persist at a tedious task for a longer duration than people lower in

  • 88 conscientiousness (Sansone, Wiebe, & Morgan, 1999).

  • 89 Agreeableness has also been linked to eVortful control ( Ahadi & Rothbart, 1994). It has

  • 90 been suggested that agreeableness is motivational in maintaining positive relationships

  • 91 with others. Indeed, it is believed that inhibitory processes are needed to control selWsh, dis-

  • 92 agreeable tendencies in group situations. For example, Havill, Besevegis, and Mouroussa-

  • 93 ski (1998) deWne agreeableness as the ability to inhibit disagreeable tendencies. Compared

  • 94 to their peers, high agreeable individuals do indeed respond to interpersonal conXict more

  • 95 constructively (Graziano, Jensen-Campbell, & Hair, 1996; Jensen-Campbell, Graziano, &

  • 96 Hair, 1996), cooperate more productively during interdependent group tasks (Graziano,

  • 97 Hair, & Finch, 1997) and are described by their parents has exhibiting more competent

  • 98 social skills (Jensen-Campbell, 2006).

  • 99 Direct links between agreeableness and eVortful control processes have also been found.

    • 100 For example, agreeableness is predictive of eVorts to control one’s emotions in both adults

    • 101 and children (Tobin & Graziano, 2006; Tobin, Graziano, & Vanman, 2000). In addition,

    • 102 Jensen-Campbell et al. (2002) found that agreeableness uniquely predicts Stroop perfor-

    • 103 mance, which is believed to be a measure of executive control ability. Finally, agreeableness

    • 104 has been linked with orienting sensitivity, which is related with associative sensitivity as

    • 105 well as sensitivity to internal, aVective, and external perception (Rothbart et al., 2001 ).

    • 106 Recent research has found that both agreeableness and conscientiousness were associ-

    • 107 ated with resistance to temptation in early adolescence (Jensen-Campbell & Graziano,

    • 108 2005). SpeciWcally, Jensen-Campbell and Graziano found that in situations with mixed

    • 109 messages about cheating, high conscientious children were less likely to cheat than were

    • 110 children low in conscientiousness. In addition, they found that when cheating was clearly

    • 111 conveyed to be “wrong,” children high in teacher-rated agreeableness cheated less than

ARTICLE IN PRESS

  • 4 L.A. Jensen-Campbell et al. / Journal of Research in Personality xxx (2006) xxx–xxx

  • 112 children low on teacher-rated agreeableness did. Finally, adolescents who were low in both

  • 113 agreeableness and conscientiousness were the most likely to cheat (see Fig. 1, Jensen-

  • 114 Campbell & Graziano, 2005).

  • 115 In addition, both agreeableness and conscientiousness have been associated with tradi-

  • 116 tional clinical assessments associated with poor regulation, namely attention deWcit hyper-

  • 117 activity disorder (ADHD). It should be noted that ADHD is believed to be a disorder that

  • 118 involves poor inhibitory control (Barkley, 2001). Using the Child Behavior Checklist

  • 119 (CBCL), parents report greater attention problems and more behaviors associated with

  • 120 ADHD for children who were low on agreeableness and conscientiousness (Jensen-Camp-

  • 121 bell, 2006). Similarly, Nigg, Blaskey, Huang-Pollock, and Rappley (2002) found that both

  • 122 conscientiousness and agreeableness were diVerentially associated with ADHD symptom

  • 123 clusters. For instance, conscientiousness predicted inattention-disorganization, whereas

  • 124 agreeableness was associated with hyperactivity-impulsivity and oppositional deWant

  • 125 behaviors.

  • 126 1.2. Neural correlates of self-control

  • 127 Self-control processes, such as emotional control and attention, have been associated

  • 128 with prefrontal cortex functioning (Coan & Allen, 2004; Goldberg, 2001). For example,

  • 129 individuals who suVer injuries to the prefrontal cortex often suVer from poorer emotional

  • 130 control (Fuster, 1989; Harmon-Jones, 2004). Numerous studies have also found an associ-

  • 131 ation between left anterior brain activity and the behavioral activation system, and right

  • 132 anterior brain activity and the behavioral inhibition system. Moreover, this brain asymme-

  • 133 try is remarkably stable across time (Tomarken, Davidson, Wheeler, & Kinney, 1992). Har-

  • 134 mon-Jones and Allen (1998) and Harmon-Jones and Sigelman (2001) have found that

  • 135 greater left prefrontal brain activity, in comparison to right prefrontal brain activity, is

  • 136 related to self-reported anger, a negatively valenced approach emotion. In addition, Har-

  • 137 mon-Jones (2004) has found evidence of a direct link between a person’s behavioral

  • 138 approach orientation and anger responses. Persons with strong approach orientations are

  • 139 more likely to experience anger than are persons with weaker approach orientations.

  • 140 Not only is frontal activation asymmetry related to emotion, but it is also related to

  • 141 actual social behavior. For example, Fox et al. (1995) have found that frontal activation

  • 142 asymmetry was related to social competence in preschool children. D’Alfonso, van Honk,

  • 143 Hermans, Postma, and de Haan (2000) also found that college women were more likely to

  • 144 attend to angry faces when there was increased activation in the left prefrontal cortex.

  • 145 Conversely, the women were less attentive to angry faces with increased right frontal acti-

  • 146 vation.

  • 147 1.3. The present study

  • 148 The link between agreeableness and conscientiousness to both prefrontal cortical activ-

  • 149 ity and emotional/behavioral regulation was examined. We were speciWcally interested in

  • 150 how individuals handle frustration that comes from receiving negative feedback. Some the-

  • 151 orists may argue that self-control is almost a deWnitional part of agreeableness and consci-

  • 152 entiousness. For example, one of the presumed facets of conscientiousness on the NEO is

  • 153 deliberation (i.e., considering the consequences before taking action; McCrae & Costa,

  • 154 1996). Other researchers (e.g., Block, 1995), however, have pointed to deWciencies in the

ARTICLE IN PRESS

L.A. Jensen-Campbell et al. / Journal of Research in Personality xxx (2006) xxx–xxx

5

155

Wve-factor approach as a general perspective on personality structure. Despite the theoreti-

156

cal links between self-regulation and personality, there has been little research that directly

157

examines these links, especially in terms of prefrontal cortex (PFC) functioning. If agree-

158

ableness and conscientiousness involve individual diVerences in executive attention and

159

emotional regulation, there should be evidence of diVerences in prefrontal cortex function-

160

ing, especially when confronted with frustrating situations.

161

This study was designed to answer two ke y empirical questions: (1) Are conscientious-

162

ness and agreeableness related to anger and aggressive responses; and (2) Do these person-

163

ality traits moderate the link between anger and aggression? We speciWcally examined the

164

role of PFC asymmetry and its association to agreeableness and conscientiousness. It was

165

anticipated that individuals lower in conscientiousness and agreeableness will exhibit more

166

left relative to right prefrontal brain asymmetry, will report more anger, and will aggress

167

more, but only when they are confronted with a situation that requires greater self-control

168

(i.e., when they receive negative feedback from another person). Second, we anticipated an

169

important buVering inXuence associated with agreeableness and conscientiousness. That is,

170

it was expected that the known link between anger and aggression would only hold for per-

171

sons low on agreeableness and conscientiousness. In other words, even if persons higher on

172

agreeableness and conscientiousness experience the same or greater levels of anger than

173

persons lower on these dimensions, the association between anger and aggression will be

174

weaker for persons higher on agreeableness and conscientiousness because of their greater

175

eVortful control capabilities.

176

This is not to say that other Big Five personality constructs, such as extraversion and neu-

177

roticism, are not important for understanding the anger–aggression link. Extraversion

178

involves surgency, sociability, and social interest (Elphick, Halverson, & Marszal-Wis-

179

niewska, 1998). Indeed, Shiner (2000) has found that extroverted children are more compe-

180

tent in social situations than their less extroverted peers are. Thus, individuals higher in

181

extraversion may be less likely to experience anger when confronted with negative feedback

182

and may be less likely to aggress. Neuroticism, on the other hand, is related to a person’s

183

emotional stability. Neurotic individuals are often easily frustrated. Indeed, they have a ten-

184

dency to be hypersensitive to negative events, which may lead to greater negative aVect when

185

confronted with negative feedback from another person (Suls, Martin, & David, 1998).

186

2. Methods

187

  • 2.1. Participants

188

Sixty-three men and 63 women participated for partial fulWllment of an Introductory

189

Psychology course requirement. The racial composition of the sample included 46.8% Cau-

190

casian, 23% Asian, 14.3% African American, 5.6% American Indian, and 10.3% other. Of

191

the participants, 15.1% indicated Hispanic or Latino ethnicity. The participants ranged in

192

age from 18.38 to 47.74 years (M D 21.24, SD D 4.03).

193

  • 2.2. Materials

194

2.2.1. Measures of personality

195

To assess personality, participants completed both the Big Five Inventory (BFI; John,

196

Donahue, & Kentle, 1991) and Goldberg’s (1992) Trait Markers. The BFI is a 44-item

ARTICLE IN PRESS

  • 6 L.A. Jensen-Campbell et al. / Journal of Research in Personality xxx (2006) xxx–xxx

  • 197 inventory consisting of short phrases based on trait adjectives associated with each of the

  • 198 Big Five personality factors (e.g., is helpful and unselWsh with others; is a reliable worker).

  • 199 Participants rated the degree to which they agreed or disagreed with each statement on a

  • 200 Likert-type scale of 1 (strongly agree) to 5 (strongly disagree). The BFI had respectable

  • 201 reliabilities ranging from .71 (extroversion) to .80 (conscientiousness) as assessed by Cron-

  • 202 bach’s . The for agreeableness was .74.

  • 203 For Goldberg’s (1992) trait markers, participants rated the degree to which they agreed

  • 204 or disagreed with 100 trait words (e.g., conscientious, agreeable) on a 5-point Likert-type

  • 205 scale of 1 (strongly disagree) to 5 (strongly agre e). Instead of presenting the traits in bipolar

  • 206 format, we separated the poles and presented them in unipolar format (i.e., 20 words for

  • 207 each factor). DiVerence scores were computed for appropriate pairs (see Briggs, 1992, p.

  • 208 268). Goldberg’s trait markers had strong reliabilities ranging from .74 (emotional stabil-

  • 209 ity) to .88 (openness to experience). The Cronbach’s s for agreeableness and conscien-

  • 210 tiousness were .80.

  • 211 Our two assessments were highly correlated for extraversion (r(126) D .81, p < .01),

  • 212 agreeableness (r(126) D .65, p < .01), conscientiousness (r(124) D .51, p > .01), emotional sta-

  • 213 bility/neuroticism (r(124) D ¡.56, p < .01), and openness to experience (r(124) D .55, p < .01).

  • 214 Given this convergence and that the focus of the paper was not to assess the convergent

  • 215 and discriminate validity among Big Five measures, we chose to collapse across our per-

  • 216 sonality assessments to make a unitary reliable assessment of each Big Five dimension. To

  • 217 create these assessments, we standardized the scores on each Big Five dimension and then

  • 218 averaged across the two assessments.

  • 219 2.2.2. Measures of emotion

  • 220 Participants rated themselves on 24 emotions at the beginning and at the end of the

  • 221 study using a Likert-type scale ranging from 1 (very slightly or not at all) to 5 (extremely).

  • 222 Emotion words included angry, bad, irritable, annoyed, agitated, hostile, frustrated, afraid,

  • 223 scared, nervous, jittery, good mood, happy, uplifted, alert, active, determined, enthusiastic,

  • 224 excited, inspired, interested, proud, strong, and attentive. The Wrst assessment served as a

  • 225 baseline measure and the second assessment served as a state measure of emotion. These

  • 226 emotion words were identical to the ones used in Harmon-Jones and Sigelman (2001) (see
    227 also Watson, Clark, & Tellegen, 1988). Only the trait word anger was used for this study. 1

  • 228 2.3. EEG recordings

  • 229 EEG acquisition followed standard guidelines (Pivik et al., 1993). An Electro-Cap was

  • 230 positioned according to the 10–20 International System (Jasper, 1958). EEG was recorded

  • 231 from mid-frontal (F3, F4), lateral-frontal (F7, F8), parietal (P3, P4), midline frontal (Fz),

  • 232 and midline parietal (Pz) sites. EEG sites were recorded in reference to the left earlobe

  • 233 (A1), and data were acquired from the right earlobe (A2) to derive digitally the averaged

  • 234 ear reference (Hagemann, 2004). The isolated-common ground was the frontal pole site

  • 235 (Fpz). The electrooculogram (EOG) was recorded with aYxed 6-mm tin electrodes to the

1 Additional analyses examined the other emotion words by creating three composites: negative aVect (angry, bad, irritable, annoyed, agitated, hostile, frustrated), fear (afraid, scared, nervous, jittery), and positive aVect (good mood, happy, uplifted, alert, active, determined, enthusiastic, excited, inspired, interested, proud, strong, at- tentive). Only negative aVect, which included angry, produced expected diVerences among groups.

ARTICLE IN PRESS

L.A. Jensen-Campbell et al. / Journal of Research in Personality xxx (2006) xxx–xxx

7

  • 236 right outer canthus and superior orbit to monitor eye movements. Scalp electrode imped-

  • 237 ances were below 5 k , and left–right electrode pairs were within 1 k of each other, result-

  • 238 ing in low electrical noise in the EEG signals.

  • 239 Using hardware and software from the James Long Company (Caroga Lake, NY), we

  • 240 collected baseline EEG data by having participants relax with the eyes open or closed for

  • 241 alternating eight 1-min intervals. Participants were randomly assigned to start the baseline

  • 242 with their eyes open or eyes closed. Data were also collected for 1 min following the labora-

  • 243 tory task. The bioampliWer was set for band-pass Wltering with half power cutoV frequen-

  • 244 cies of 1 and 200 Hz (12 dB/octave rollo V). The gain was 10,000 for EEG channels and 2500

  • 245 for EOG channels. Data were digitized continuously at 1024 Hz.

  • 246 2.4. Procedure

  • 247 Participants came into the laboratory twice. In the Wrst session, they completed self-

  • 248 report measures. To assess the Big Five, participants completed the BFI (John et al., 1991)

  • 249 and Goldberg’s (1992) trait markers. Other measures were also collected for use in another

  • 250 study, which helped disguise the true purpose of this study. At the end of the Wrst session,

  • 251 interested right-handed participants were scheduled individually for the second session of

  • 252 the study.

  • 253 Several weeks later, participants were brought back to the lab. Following procedures

  • 254 outlined by Harmon-Jones and Sigelman (2001), investigators told participants that they

  • 255 would be involved in a perception study that would include both a person perception and a

  • 256 taste perception segment. They were then told that there was another participant in an

  • 257 adjacent room and that they would be working together on the perception tasks. In reality,

  • 258 there was no second participant. The participant then completed an emotional assessment

  • 259 questionnaire that served as a baseline measure.

  • 260 Electrodes were then attached to the participant’s head, face, and ears following stan-

  • 261 dard guidelines (Pivik et al., 1993). Participants were then told to relax, so the investigator

  • 262 could obtain baseline measures of their brain’s electrical activity. Baseline electroencepha-

  • 263 lographic (EEG) activity was recorded for eight 1-min alternating intervals with eyes open

  • 264 and eyes closed. The participant was randomly assigned to begin baseline with either eyes

  • 265 open or eyes closed. During the baseline recordings, participants were told to look straight

  • 266 ahead while trying not to move their head and body during the recordings.

  • 267 The participant was then told that he/she would be randomly assigned either to write an

  • 268 essay on an important social issue or to give his/her perception of a person who wrote such

  • 269 an essay. In fact, all participants were assigned to the role of writing the essay, which

  • 270 involved writing a 5-min essay taking a stance (for/against) on a topic of their choice. The

  • 271 topic choices were the legality of smoking in public places, a reduction in the legal drinking

  • 272 age, increasing gun control measures, or an increase in tuition to support a campus beauti-

  • 273 Wcation campaign. Participants were instructed to choose the topic that was most impor-

  • 274 tant to them personally.

  • 275 After the participant completed the essay, the investigator collected the envelope and

  • 276 delivered it to a Wctitious “other” participant for evaluation. Individuals were randomly

  • 277 assigned to receive either relatively positive or negative feedback on their essay from the

  • 278 supposed other participant. The evaluation consisted of Likert-type ratings made by the

  • 279 supposed other participant on several 9-point bipolar scales that gave his/her perception of

  • 280 the actual participant, which included unintelligent–intelligent, thought-provoking–boring,

ARTICLE IN PRESS

  • 8 L.A. Jensen-Campbell et al. / Journal of Research in Personality xxx (2006) xxx–xxx

  • 281 friendly–unfriendly, illogical–logical, respectable–unrespectable, irrational–rational. For

  • 282 the relatively negative rating, participants received ratings of 2 or 3 for assessments in

  • 283 which 9 represented a positive rating; they received ratings of 7 for assessments in which

  • 284 the positive words represented a 1. The relatively positive evaluation reversed these ratings.

  • 285 In addition, participants received an additional handwritten comment from the alleged

  • 286 other participant that corresponded with the positive or negative ratings (i.e., “I can under-

  • 287 stand why a person would think like this. This person has obviously learned something

  • 288 while at UTA.” or “I cannot understand why a person would think like this. I hope this

  • 289 person learns something while at UTA.”).

  • 290 The experimenter then returned and gave the participant their essay and the ratings/

  • 291 comments from the Wctitious “other” participant. Once the participants read their ratings

  • 292 and comments, an additional 1 min of EEG activity was recorded to measure the physio-

  • 293 logical reaction. The participants were then told that they would start the second percep-

  • 294 tion task (i.e., taste perception task), in which they would be randomly assigned to either

  • 295 choose a drink that the other participant would have to consume in its entirety or to con-

  • 296 sume a substance that the other participant had chosen for them and evaluate it on a list of

  • 297 given characteristics.

  • 298 Participants were told that it was very important for the experimenter to remain blind

  • 299 to the type of taste to which the participant was exposed. One way to keep the experi-

  • 300 menter blind to the taste is to have one participant assign the taste to the other partici-

  • 301 pant. The investigator then checked their “randomization sheet” and stated that it

  • 302 looked like the participant had been assigned to determine the drink choice. In actuality,

  • 303 all participants were assigned to choose the taste for the “other” participant. The

  • 304 researcher then left the room to prepare the “other” participant and to get the drink tray.

  • 305 When the researcher returned with the drink tray, the participant was told that he/she

  • 306 would choose between six drinks that the “other” participant had ranked from a longer

  • 307 list of substances in order of preference. Each drink consists of 11 oz. of water with 1, 2,

  • 308 or 3 teaspoons of sugar, apple juice, lemon juice, salt, vinegar, or hot sauce mixed into

  • 309 the water.

  • 310 Participants were to choose a substance and then pour some of each of the three con-

  • 311 centrations into one of the three cups provided. They were then told to cover the unused

  • 312 beverage choices with a cloth, so the experimenter would be blind to the drink choice. The

  • 313 choice of beverage signiWed the level of aggression that the participant chose to engage in.

  • 314 In other words, each substance was given a value ranging from 1 (most pleasant) to 6 (most

  • 315 aversive). Finally, the participant completed a second emotional assessment. Electrodes

  • 316 were then removed and the participant was debriefed.

  • 317 2.5. EEG analysis

  • 318 EEG was used to assess left (relative to right) prefrontal brain asymmetry and left (rela-

  • 319 tive to right) parietal brain asymmetry. The EEG data were visually inspected to identify

  • 320 epochs containing ocular and other artifacts. Using the most commonly used method to

  • 321 remove EEG artifact (Hagemann, 2004), we excluded artifact-contaminated epochs from

  • 322 the analysis. An average of 98.54 artifact-free epochs for each minute were obtained in the

  • 323 8-min baseline EEG recording, in addition to an average 88.53 artifact-free epochs that

  • 324 were obtained from the 1-min experimental EEG recording. The artifact-free epochs for

  • 325 each participant were at least greater than 13 artifact-free epochs.

ARTICLE IN PRESS

L.A. Jensen-Campbell et al. / Journal of Research in Personality xxx (2006) xxx–xxx

9

  • 326 Data were digitally transformed from an A1-reference to an average-ears reference, and

  • 327 then the Fast Fourier Transformation (FFT) was applied to each 1-min recording period.

  • 328 Hanning windows identiWed 1.00-s epochs of artifact-free data in each recording period,

  • 329 and epochs were overlapped by 50%. Spectral power from 1 Hz bins was clustered into

  • 330 broad bands. The band of primary interest was alpha (7.5–13.5 Hz), given that previous

  • 331 asymmetry and emotion results pertain speciWcally to this band. Alpha power (pW ) was

  • 332 averaged across baseline. Asymmetry metrics were computed as the ratio of right alpha

  • 333 power to left alpha power (Hagemann, 2004). 2 Given that alpha power varies inversely

  • 334 with cortical activation, alpha suppression reXects increased brain activity, and higher

  • 335 asymmetry values represent greater left relative to right activation.

  • 336 3. Results

  • 337 First, we anticipated that persons in the negative feedback condition would report more

  • 338 anger, show greater left relative to right prefrontal asymmetry, and show greater levels of

  • 339 aggression (i.e., more negative drink choices) than persons in the positive feedback condi-

  • 340 tion (Hypothesis 1a). Next, we examined the correlations among our measures of anger

  • 341 and aggression (Hypothesis 1b). It was anticipated that self-reported anger would be posi-

  • 342 tively associated with prefrontal asymmetry in the post-experimental conditions, but not

  • 343 with baseline or parietal measures. It was also anticipated that self-reported anger and pre-

  • 344 frontal asymmetry would be positively correlated with drink choice, especially in the nega-

  • 345 tive feedback condition.

  • 346 Second, we examined whether agreeableness and conscientiousness predicted reports of

  • 347 anger in the negative feedback condition (Hypothesis 2a). Anger was assessed with both

  • 348 self-report measures and prefrontal brain asymmetry. It was anticipated that conscientious

  • 349 and agreeable participants would report less anger and show less left relative to right pre-

  • 350 frontal brain asymmetry than persons lower on conscientiousness and agreeableness do

  • 351 when they receive negative feedback. In addition, it was anticipated that conscientiousness

  • 352 and agreeableness would predict aversive drink choice (Hypothesis 2b).

  • 353 Finally, we examined whether conscientiousness and agreeableness moderated the link

  • 354 between anger and aversive drink choice (Hypothesis 3). It was anticipated that the anger–

  • 355 aggression link would hold only for persons lower on Big Five personality traits associated

  • 356 with self-control. In other words, it was anticipated that anger would only predict drink

  • 357 choice for individuals who were lower on agreeableness and conscientiousness.

  • 358 Hypothesis 1a. To assess whether feedback inXuenced anger, we conducted two 2 (positive

  • 359 vs. negative feedback) £ 2 (sex of participant) ANOVAs. The dependent measures were

  • 360 self-reported baseline and post-experimental measures of anger, mid-frontal asymmetry,

  • 361 lateral-frontal asymmetry, parietal asymmetry, as well as beverage drink choice. 3

  • 2 Typically, natural log transformations are taken for each EEG site to normalize data. Taking the natural log of each site created unacceptable levels of kurtosis (>30) and skewness (>¡3.77). Using our raw data, we created ratiometric scores (e.g., F4 alpha power/F3 alpha power) to assess asymmetry while controlling for individual diVerences in skull morphology.

  • 3 Supplementary analyses were conducted in which either baseline anger or baseline asymmetry was used as a covariate in the analysis (i.e., ANCOVA). The results held for self-reported anger; the pattern of results was simi- lar, but nonsigniWcant for lateral-frontal asymmetry.

ARTICLE IN PRESS

  • 10 L.A. Jensen-Campbell et al. / Journal of Research in Personality xxx (2006) xxx–xxx

  • 362 We found that persons who received negative feedback were more angry (M D 2.02,

  • 363 SD D 1.10) than persons in the positive feedback condition (M D 1.25, SD D .63),

  • 364 F(1,121) D 28.98, p < .001, 2 D .19. There was no signiWcant diVerence between our experi-

  • 365 mental groups on baseline measures of self-reported anger, F(1,122) D .39, ns, 2 D .003

  • 366 (M D 1.39, 1.34, SD D .82, .71, for negative and positive conditions, respectively).

  • 367 Participants in the negative feedback condition showed greater left lateral-frontal activ-

  • 368 ity (M D 1.09, SD D .19) than did persons in the positive feedback condition (M D 1.02,

  • 369 SD D .23), F(1,120) D 3.04, p < .05, one-tailed, 2 D .025. There were no signiWcant diVerences

  • 370 between experimental groups for baseline measures of lateral-frontal activity,

  • 371 F(1,120) D 2.19, ns, 2 D .018 (M D 1.05, 1.00, SD D .19, .17, for negative and positive condi-

  • 372 tions, respectively).

  • 373 Contrary to our predictions, however, there were no signiWcant diVerences between

  • 374 experimental conditions on left relative to right mid-frontal activity, F(1,120) D .02, ns,

  • 375 2 D .000. As anticipated, participants in the negative feedback condition showed no

  • 376 greater left parietal activity (M D 1.06, SD D .23) than did persons in the positive feedback

  • 377 condition (M D 1.08, SD D .31), F(1,120) D .01, ns, 2 D .000. There were also no diVerences

  • 378 between groups on baseline measures of mid-frontal and parietal activity F(1,120) D .22,

  • 379 1.84, p D .64, .18, respectively.

  • 380 Finally, we examined whether the experimental condition inXuenced aversive drink

  • 381 choice. Results showed that persons in the negative feedback condition chose more aver-

  • 382 sive drinks for the other participant (M D 4.03, SD D 1.73) than did persons who were in

  • 383 the positive feedback condition (M D 2.58, SD D 1.21), F(1,122) D 28.63, p < .001, 2 D .19.

  • 384 There were no main eVects or interactions associated with the sex of the participant for any

  • 385 of our anger or aggression measures.

  • 386 Hypothesis 1b. Next, we examined the correlations among our baseline and post-experi-

  • 387 mental measures (see Tables 1 and 2). Self-reported anger was associated with left relative

  • 388 to right prefrontal asymmetry, r(124) D .22, .16, p < .02, .08 (for mid-frontal and lateral-

  • 389 frontal, respectively). The relation between mid-frontal and self-reported anger only held

  • 390 in the negative feedback condition (see Table 2; r D .34, p < .01). As anticipated, there was

  • 391 no evidence that parietal asymmetry was related to self-reported anger across conditions, Table 1 Correlations among baseline and post-experimental measures

Post-experimental measures

Beverage choice

Anger

Mid-frontal

Lateral-frontal

Parietal

Anger (baseline)

Post-experimental measures

Anger

Mid-frontal

Lateral-frontal

Parietal asymmetry

0.33

¤¤

0.04

0.18 ¤

¡0.01

Baseline measures

Anger

Mid-frontal

Lateral-frontal

Parietal asymmetry

0.09

¡0.08

0.01

¡0.12

0.22

¤

0.16

9

0.03

0.30

¤¤

0.11

0.53 ¤¤

0.05

¡0.02

¡0.01

0.08

0.65

¤¤

0.14

0.21 ¤

0.06

0.12

0.23

¤

0.62 ¤¤

0.05

0.07

0.06

0.01

0.50 ¤¤

¡0.12

¡0.05

¡0.01

Note: ¤ p < .05; ¤¤ p < .01; 9 p < .05, one-tailed.

ARTICLE IN PRESS

L.A. Jensen-Campbell et al. / Journal of Research in Personality xxx (2006) xxx–xxx

11

Table 2 Correlations among post-experimental measures by type of feedback

Post experimental measures

Beverage

choice

Anger

Mid-frontal

asymmetry

Lateral-frontal

asymmetry

Negative feedback (N D 60) Anger Mid-frontal asymmetry Lateral-frontal asymmetry Parietal asymmetry

Positive feedback (N D 64) Anger Mid-frontal asymmetry Lateral-frontal asymmetry Parietal asymmetry

0.31

¤¤

0.14

0.21

9

0.10

¡0.07

¡0.16

0.05

¡0.10

0.34

¤¤

0.08

0.24

9

0.03

0.16

¡0.19

0.27

¤¤

0.19

0.35 ¤¤

0.05

0.20

¡0.02

Note: ¤¤ p < .01; 9 p < .05, one-tailed.

  • 392 r(124) D .03, ns. Self-reported anger was associated with aversive drink choice (r(126) D .33,

  • 393 p < .01) and this association held only for the negative feedback condition (r(60) D .31,

  • 394 p < .01). Baseline measures of anger were not associated with each other or with the aggres-

  • 395 sive behavior of the participant, r < .13 (see Table 1). However, baseline self-reported

  • 396 anger was signiWcantly correlated with post-experimental self-reported anger, r(126) D .53,

  • 397 p < .01. In addition, mid-frontal and lateral-frontal asymmetry were highly correlated from

  • 398 pre- to post-observations, r(123) D .65, .62, p < .01, respectively. Likewise, parietal asymme-

  • 399 try was highly correlated from pre- to post-observations, r(123) D .50, p < .01.

  • 400 Overall, the pattern of results suggests that self-reported anger was associated with pre-

  • 401 frontal asymmetry. First, participants in the negative feedback condition reported more

  • 402 anger and showed greater lateral-frontal asymmetry. Second, self-reported anger was also

  • 403 associated with both mid-frontal and lateral-frontal asymmetry across conditions. In addi-

  • 404 tion, mid-frontal asymmetry was associated with anger in the negative feedback condition,

  • 405 but not in the positive feedback condition. As previous research has found, there was also

  • 406 stability between our baseline and post-experimental ratings of anger. Finally, type of feed-

  • 407 back predicted aversive drink choice; persons in the negative feedback condition chose

  • 408 more aversive drinks. Moreover, anger and lateral-frontal asymmetry predicted drink

  • 409 choice in the negative feedback condition.

  • 410 Hypothesis 2. The main purpose of this research, however, was to determine if Big Five person-

  • 411 ality dimensions associated with self-control inXuence emotional and behavioral responses

  • 412 when individuals are confronted with a frustrating situation. Subsequent analyses were run

  • 413 separately by condition because of power issues associated with detecting three-way interac-

  • 414 tions (see Aiken & West, 1991, p. 139). Hierarchical regression analyses were performed to

  • 415 determine the unique predictiveness of agreeableness and conscientiousness. Extroversion,

  • 416 emotional stability, and openness to experience were centered and entered on the Wrst step as

  • 417 control variables; agreeableness, conscientiousness, and their cross-product was then entered

  • 418 on the second step. Criterion measures included drink choice as well as baseline and post-levels

  • 419 of anger, mid-frontal asymmetry, lateral-frontal asymmetry, and parietal asymmetry.

  • 420 For self-reported anger, agreeableness, conscientiousness, and their cross-product pro-

  • 421 duced a signiWcant increase in variance explained over the other three Big Five dimensions in

ARTICLE IN PRESS

  • 12 L.A. Jensen-Campbell et al. / Journal of Research in Personality xxx (2006) xxx–xxx

  • 422 the negative feedback condition, F (3,54) D 4.67, p < .01, R 2 D .20. Conscientiousness was

  • 423 uniquely and negatively related to self-reported anger B .51, t(54) 2.44, p <.02,

  • 424 sr 2 D .08. Agreeableness, however, was not related to self-reports of anger in the predicted

  • 425 direction, B D .39, t(54) D 1.96, p <.06 (see Table 3). There was, however, an

  • 426 agreeableness £ conscientiousness interaction for self-reported anger in the negative feedback

  • 427 condition, t(54) 2.50, ¡2.88, p <.02, sr 2 D .09. Using procedures outlined by Aiken and

  • 428 West (1991), we found that agreeableness was positively related to self-reported anger when

  • 429 conscientiousness was low (¡1 SD), B D .71, sr 2 D .14, ts(54) D 3.21, p < .01. When conscien-

  • 430 tiousness was high, there was no evidence that agreeableness was associated with self-

  • 431 reported anger, B D .07, ts(54) D .29, sr 2 D .001, ns. There was no evidence that agreeableness

  • 432 and conscientiousness predicted self-reported anger in the positive feedback condition

  • 433 (p > .10) (see Table 3).

  • 434 For lateral-frontal asymmetry, agreeableness, conscientiousness, and their cross-prod-

  • 435 uct also produced a signiWcant increase in variance explained over the other three Big Five

  • 436 dimensions in the negative feedback condition, F (3,53) D 2.70, p < .01, R 2 D .12. Consci-

  • 437 entiousness was related to lateral-frontal activity in the predicted direction, B D ¡.06,

  • 438 sr 2 D .04, t(53) D ¡1.76, p < .05, one-tailed. There was again no evidence that agreeableness

  • 439 was related to lateral-frontal activity in the predicted direction, t(53) D 1.45, p D .15. There

  • 440 was an agreeableness £ conscientiousness interaction for lateral-frontal activity in the neg-

  • 441 ative feedback condition, t(53) D ¡1.99, p < .05, sr 2 D .06. Again, we found that agreeable-

  • 442 ness was positively related to prefrontal asymmetry when conscientiousness was low (¡1

  • 443 SD), B D .10 sr 2 D .09, t(53) D 2.45, p < .02. When conscientiousness was high, there was no

  • 444 evidence that agreeableness was associated with prefrontal asymmetry, B D .01, t(53) D .14,

  • 445 sr 2 D .0002, ns. There was no evidence that agreeableness and conscientiousness predicted

  • 446 lateral-frontal activity in the positive feedback condition (p > .10) (see Table 3). Table 3 B-weights and semi-partial correlations: the Big Five dimensions of personality predicting post-experimental levels of anger and aggression

 

Anger

Lateral-frontal

Mid-frontal

Parietal

Beverage

 

asymmetry

asymmetry

asymmetry

choice

Negative condition (N D 59) Extroversion Emotional stability Openness to experience Conscientiousness Agreeableness Conscientiousness £ agreeableness

Positive condition (N D 64)

0.13 (0.09) ¡0.31 (¡0.20) 0.05 (0.03) ¡0.51 (¡0.29) ¤ 0.39 (0.23) ¡0.37 (¡0.30) ¤

0.00 (0.01) 0.02 (0.06) 0.08 (0.30) ¤ ¡0.06 (¡0.21) 9 0.05 (0.18) ¡0.05 (¡0.24) ¤

0.01(0.07)

¡0.06 (¡0.29) ¤ 0.05 (0.23) ¡0.03 (¡0.12) 0.04 (0.19) ¡0.01 (¡0.06)

0.07 (0.23) 0.02 (0.06) ¡0.02 (¡0.07) ¡0.05 (¡0.14) 0.09 (0.26) ¤ 0.02 (0.07)

0.60 (0.27) ¤ 0.04 (0.02) ¡0.10 (¡0.04) 0.12 (0.04) 0.19 (0.07) ¡0.17 (¡0.09)

Extroversion

0.08 (0.10)

0.07 (0.25) ¤

0.01 (0.06)

¡0.04 (¡0.11)

¡0.03 (¡0.02)

Emotional stability

¡0.26 (¡0.30) ¤

¡0.00 (¡0.00)

0.02 (0.14)

0.08 (0.20)

¡0.15 (¡0.09)

Openness to experience

0.13 (0.17)

¡0.08 (¡0.29) ¤

¡0.05 (¡0.35) ¤¤

¡0.06 (¡0.16)

0.29 (0.19)

Conscientiousness

0.12 (0.12)

0.03 (0.09)

0.02 (0.10)

0.04 (0.08)

¡0.12 (¡0.06)

Agreeableness

¡0.13 (¡0.16)

¡0.04 (¡0.13)

¡0.01 (¡0.07)

¡0.08 (¡0.20)

0.02 (0.01)

Conscientiousness £ agreeableness

0.10 (0.11)

¡0.01 (¡0.03)

¡0.04 (¡0.21)

¡0.03 (¡0.07)

0.14 (0.08)

Note: srs are in parentheses; ¤ < .05; ¤¤ p < .01; 9 p < .04, one-tailed direction prediction.

ARTICLE IN PRESS

L.A. Jensen-Campbell et al. / Journal of Research in Personality xxx (2006) xxx–xxx

13

  • 447 As expected, there was no evidence that conscientiousness was systematically related to

  • 448 parietal asymmetry in the negative feedback condition, t(53) D ¡1.14, p D .26. However,

  • 449 agreeableness was positively associated with parietal activity, t(53) D 2.06, p < .05. Neither

  • 450 agreeableness nor conscientiousness predicted parietal activity in the positive feedback

  • 451 condition, t(57) D ¡1.61, .67, p D .11, .51, respectively. Conscientiousness did not systemati-

  • 452 cally predict outcomes for baseline measures of anger in the negative feedback condition,

  • 453 t(53) D ¡1.15. ¡.54, ¡1.62, p D .25, .59, .11 for self-reported anger, mid-frontal asymmetry,

  • 454 lateral-frontal asymmetry, respectively. Similarly, agreeableness was not signiWcantly asso-

  • 455 ciated with baseline measures of anger in the negative feedback condition, t < .74, ns.

  • 456 Next, we conducted supplementary analyses to examine the contribution of each hemi-

  • 457 sphere separately in the lateral-frontal region. The asymmetry measures are useful to

  • 458 understand diVerences in alpha power while controlling for skull thickness. However, they

  • 459 do not allow us to understand which hemisphere is contributing to the asymmetrical diVer-

  • 460 ences. We wanted to evaluate further whether conscientious individuals had more activa-

  • 461 tion in their right hemisphere (i.e., the hemisphere associated with avoidant emotions) or

  • 462 whether they simply had less activation in their left hemisphere (i.e., the hemisphere associ-

  • 463 ated with approach emotions) when confronted with a frustrating situation.

  • 464 To assess the eVects of each hemisphere separately while controlling for skull morphol-

  • 465 ogy, we followed procedures outline by Wheeler, Davidson, and Tomarken (1993) and

  • 466 Harmon-Jones, Vaughn-Scott, Mohr, Sigelman, and Harmon-Jones (2004). The alpha

  • 467 power at one site (e.g., F7) was regressed onto the alpha power at all sites and the alpha

  • 468 power at the homologous site (e.g., F8). The residual was then used as our new dependent

  • 469 measure. For these analyses, it is important to understand that an increase in alpha power

  • 470 is conversely related to cortical activity.

  • 471 For left lateral-frontal activity, conscientiousness was negatively associated with corti-

  • 472 cal activity (B D .91, sr 2 D .09, t(53) D 2.48, p < .02). On the other hand, agreeableness was

  • 473 positively associated with left lateral-frontal activity in the negative feedback condition

  • 474 (B D ¡.73, sr 2 D .06, t(53) D ¡2.06, p < .04). However, there was an agreeableness

  • 475 £ conscientiousness interaction, t(53) D 2.56, p < .01. When conscientiousness was low,

  • 476 agreeableness predicted left lateral-frontal activity, B D ¡1.31, sr 2 D .16, t(53) D ¡3.33,

  • 477 p < .02. When conscientiousness was high, agreeableness was not related to left lateral-fron-

  • 478 tal activity, B D ¡.15, t(53) D ¡.35, ns.

  • 479 For right lateral-frontal activity, conscientiousness was positively associated with corti-

  • 480 cal activity in the negative feedback condition (B D ¡1.02, sr 2 D .07, t(53) D ¡2.16, p < .04).

  • 481 There was also an agreeableness £ conscientiousness interaction, t(53) D ¡1.94, p < .057.

  • 482 When conscientiousness was low, agreeableness was negatively related to right lateral-

  • 483 frontal activity, B D .94, sr 2 D .05, t(53) D 1.86, p < .05, one-tailed. When conscientiousness

  • 484 was high, this relationship did not hold, B D ¡.19, t(53) D ¡.33, ns. There was no evidence

  • 485 of lateral prefrontal hemispheric diVerences for the other Big Five dimensions or in the

  • 486 positive feedback condition. In sum, it appears that both the right and left hemispheres

  • 487 were contributing to individual diVerences in prefrontal asymmetry.

  • 488 Hypothesis 2b. Next, we examined whether personality was directly associated with aggres-

  • 489 sive behavior. Unexpectedly, we found no evidence that agreeableness and conscientious-

  • 490 ness inXuenced aversive drink choice in the negative feedback condition. First, there was

  • 491 no appreciable change in R 2 when agreeableness, conscientiousness, and their cross-prod-

  • 492 uct were added to the equation, F (3,54) D .49, ns, R 2 D .02. Neither agreeableness nor

ARTICLE IN PRESS

  • 14 L.A. Jensen-Campbell et al. / Journal of Research in Personality xxx (2006) xxx–xxx

  • 493 conscientiousness was uniquely associated with aversive drink choice, t(54) D .56, .34, ns

  • 494 (see Table 3 for B-weights). However, extraversion was positively associated with aversive

  • 495 drink choice. Persons higher on extraversion chose more aversive drinks in the negative

  • 496 feedback condition, t(54) D 2.06, p < .05.

  • 497 In sum, conscientiousness was negatively associated with two of the three post-experi-

  • 498 mental measures of anger in the negative feedback condition. Agreeableness was positively

  • 499 associated with two of the three post-experimental measures of anger, but only when con-

  • 500 scientiousness was low and when negative feedback was provided. As anticipated, neither

  • 501 agreeableness nor conscientiousness predicted anger during baseline or when receiving

  • 502 positive feedback. Contrary to our expectations, however, neither agreeableness nor con-

  • 503 scientiousness was directly related to aversive drink choice.

  • 504 There was no consistent pattern for the other Big Five dimensions (see Table 3). For

  • 505 example, openness to experience was associated with prefrontal asymmetry but not self-

  • 506 reported anger in the positive feedback condition. Emotional stability was negatively asso-

  • 507 ciated with self-reported anger but not measures of prefrontal asymmetry in the positive

  • 508 feedback condition. Conversely, emotional stability was negative related to mid-frontal

  • 509 asymmetry in the negative feedback condition, but not with lateral-frontal asymmetry or

  • 510 self-reported anger. Finally, extraversion was related to drink choice but not measures of

  • 511 anger in the negative feedback condition.

  • 512 Hypothesis 3. Next, we examined whether conscientiousness and agreeableness moderated

  • 513 the link between anger and aggression separately for each condition. To control for base-

  • 514 line levels in these analyses, we created new measures of anger and prefrontal asymmetry.

  • 515 We regressed the baseline score from the post-experimental score (Appelbaum & McCall,

  • 516 1983). The standardized residual scores were then used as a new measure of post-experi-

  • 517 mental emotion that controlled for baseline levels. Using moderated multiple regression,

  • 518 four of the Big Five dimensions were entered on the Wrst step as control variables. Either

  • 519 agreeableness or conscientiousness and self-reported anger (controlling for baseline) were

  • 520 entered on the second step; and the cross-product was entered on the Wnal step. 4

  • 521 As anticipated, there was a conscientiousness £ anger interaction in the negative feed-

  • 522 back condition, t(53) D ¡2.12, p < .04, sr 2 D .08. When persons were lower on conscientious-

  • 523 ness, anger was associated with aversive drink choice. When persons were higher on

  • 524 conscientiousness, there was no association between anger and aversive drink choice (see

  • 525 Table 4). Although weaker and not statistically signiWcant, there was a similar pattern of

  • 526 eVects for conscientiousness £ mid-frontal asymmetry (controlling for baseline) in the neg-

  • 527 ative feedback condition, B D ¡.40, t(51) D ¡1.56, p D .12, sr 2 D .02. Again, when persons

  • 528 were lower on conscientious, mid-frontal activity was associated with aversive drink choice

  • 529 (see Table 4). As anticipated, the Wnding that conscientiousness moderated the anger–

  • 530 aggression link held only for the negative feedback condition.

4 Additional moderated multiple regression models were run that examined other possible models. One model examined the possible three-way interaction between agreeableness, conscientiousness, and anger predicting ag- gression. There was no evidence of any signiW cant three-way interactions. Second, the possible interactive inXu- ence of extraversion and emotional stability was examined. There was no evidence that either extroversion or emotional stability moderated the association between anger and aggression. Moreover, there was no evidence that these dimensions moderated the conscientiousness £ anger or agreeableness £ anger interactions.

ARTICLE IN PRESS

L.A. Jensen-Campbell et al. / Journal of Research in Personality xxx (2006) xxx–xxx

15

  • 531 There was no evidence that agreeableness moderated the anger–aggression link

  • 532 (t(53) D ¡1.20, ¡1.28, p D .23, .21 for self-reported anger and mid-frontal asymmetry, respec-

  • 533 tively). However, the associations were in the predicted direction (see Table 4). In other

  • 534 words, the slope of the line between anger–aggression and mid-frontal asymmetry-aggression

  • 535 were signiWcantly diVerent from zero, but only when persons reported low or medium levels

  • 536 of agreeableness. However, there was no evidence that the slopes were signiWcantly diVerent

  • 537 from one another across levels of agreeableness (i.e., the cross-product was not signiWcant).

  • 538 To examine the moderating inXuence of conscientiousness further, we divided partici-

  • 539 pants into high and low conscientious types using median split procedures. After receiving

  • 540 negative feedback, there were no diVerences on aggression levels between high conscien-

  • 541 tious groups (M D 4.33) and low conscientious groups (M D 3.68), t(59) D ¡.87, ns. There

  • 542 were also no diVerences on variances between the high and low conscientious groups

  • 543 (SD D 1.78, 1.63, respectively), F(1,59) D 1.07, ns.

  • 544 Using t-tests for independent correlations, we found that the link between self-reported

  • 545 anger (controlling for baseline) and aggression was strongest for individuals low on consci-

  • 546 entiousness when they received negative feedback, r(28) D .58, p < .01 (see Table 5). Fur-

  • 547 thermore, this link held for individuals low on conscientiousness but not for those high in

  • 548 conscientiousness (r D .58

vs.

.01, respectively). In other words, the level of self-reported

  • 549 anger was not associated with aversive drink choice for persons high on conscientiousness,

  • 550 whereas the association was strong and signiWcant for individuals low on conscientious-

  • 551 ness. As expected, when low-conscientious individuals reported being angry, there was a

  • 552 corresponding asymmetry in the mid-frontal region as compared with high-conscientious

  • 553 individuals (r D .46 vs. ¡.01, respectively). In addition, mid-frontal and lateral-frontal

  • 554 asymmetry were positively related to beverage choice for low conscientious individuals Table 4 The relation (B-weights) of anger to aggression as a function of conscientiousness and agreeableness

¡1 SD

0 SD

+1 SD

t-value

Level of conscientiousness

Negative feedback condition Anger Mid-frontal Lateral-frontal

Positive feedback condition Anger Mid-frontal Lateral-frontal

Level of agreeableness

1.34

¤¤

0.80

¤¤

0.55

0.12

¡0.03

0.03

Negative feedback condition

Anger

Mid-frontal

Lateral-frontal

Positive feedback condition Anger Mid-frontal Lateral-frontal

1.12

¤¤

0.78 ¤

0.39

¡0.15

¡0.09

0.12

0.69

¤

0.45

¤

0.42

0.01

0.13

0.26

0.78

¤¤

0.50 ¤

0.45

0.12

0.01

0.22

0.01

0.11

0.28

0.01

0.29

0.49 ¤

0.43

0.21

0.52

0.39

0.10

0.33

¡2.12 ¤

¡1.56

¡0.47

0.96

1.24

2.01

¡1.20

¡1.28

0.28

0.56

0.03

0.67

Note: ¤ p < .05; ¤¤ p < .01.

ARTICLE IN PRESS

  • 16 L.A. Jensen-Campbell et al. / Journal of Research in Personality xxx (2006) xxx–xxx

Table 5 Comparison of correlations between low- and high-conscientious groups

Low conscientious

 

Anger

Beverage choice

28

0.58 a ¤¤

0.46

a ¤

0.57

a ¤¤

0.27

a

0.36

9

¡0.05

0.17

32

0.08

0.08

¡0.19

0.20

¡0.04

¡0.01

¡0.07

High conscientious

 

Anger

Beverage choice

33

0.01

b

¡0.01

b

0.04

b

¡0.24

b

0.06

0.07

¡0.08

33

¡0.02

¡0.10

¡0.05

0.13

0.19

¡0.18

¡0.10

Negative feedback

N

Anger Mid-frontal asymmetry Lateral-frontal asymmetry Parietal asymmetry

Positive feedback

N

Anger Mid-frontal asymmetry Lateral-frontal asymmetry Parietal asymmetry

Correlations with diVerent subscripts were signiWcantly diVerent from each other between the two conscientious groups, p < .05, one-tailed. p < .05. p < .01.

¤

¤¤

  • 9 p < .05, one-tailed.

  • 555 (r D .57, .36); this association did not hold for high conscientious individuals (r D .04). As

  • 556 expected, no signiWcant diVerences were found within the positive feedback condition.

  • 557 4. Discussion

  • 558 The current study examined the role of conscientiousness and agreeableness in inXuenc-

  • 559 ing self-control during interpersonally frustrating situations. More speciWcally, we exam-

  • 560 ined whether conscientiousness and agreeableness inXuenced anger reactions and

  • 561 aggressive responding. This study found that conscientiousness was related to both self-

  • 562 reported anger and frontal cortical asymmetry associated with anger responses (Harmon-

  • 563 Jones & Sigelman, 2001). In addition, conscientiousness moderated the association

  • 564 between anger and aggression. Individuals who were low on conscientiousness were more

  • 565 likely to chose an aversive drink when they reported being angry. There was no evidence of

  • 566 a relation between anger and aggression when individuals were high on conscientiousness,

  • 567 suggesting that individuals higher on conscientiousness may be better able to control their

  • 568 behavior when they were frustrated.

  • 569 Our Wndings suggest that conscientiousness is not only important for task performance

  • 570 (e.g., Barrick & Mount, 1991), but it may also be critical for individuals to be responsive to

  • 571 social self-regulation rules. Traits associated with self-control may have evolved speciWcally

  • 572 for reciprocal social exchanges and for the formation of cooperative group living (Barkley,

  • 573 2001). Other researchers have posited these ideas as well (see Baumeister & Leary, 1995; Bau-

  • 574 meister & Vohs, 2004; Rothbart & Bates, 1998). Indeed, poor general self-regulation has been

  • 575 linked to poor emotional control (Barkley, 1998; Baumeister, Heatherton, & Tice, 1994; Bau-

  • 576 meister & Vohs, 2004). In other words, conscientiousness may be important for interpersonal

  • 577 relationships because it serves as an indication of executive control functions such as behav-

  • 578 ioral inhibition (Nigg, 2000), which in turn is important for developing and maintaining inter-

  • 579 personal relationships. Indeed, Jensen-Campbell and Malcolm (2006) have found that

ARTICLE IN PRESS

L.A. Jensen-Campbell et al. / Journal of Research in Personality xxx (2006) xxx–xxx

17

  • 580 adolescents who are lower in conscientiousness seem to suVer in their peer relationships. This

  • 581 study suggests one of the possible routes for poorer interpersonal relationships may be

  • 582 because low conscientious individuals may be less able to control their anger when con-

  • 583 fronted with a frustrating interpersonal situation. Indeed, emotional and behavioral regula-

  • 584 tion has been linked to social competence and peer conXicts in children as young as preschool

  • 585 age (2–5 years old). Socially competent children are those who have and exercise the ability to

  • 586 adjust their behavior appropriately across situations (Cavell, 1990).

  • 587 Agreeableness, however, was positively associated with anger, but not aggressive behav-

  • 588 ior in the negative feedback condition. One possible explanation for this result is that per-

  • 589 sons higher in agreeableness may display more negative aVect than persons lower in

  • 590 agreeableness because receiving negative feedback may represent a greater mismatch of

  • 591 their interpersonal orientation. The idea is that people react best to situations that best Wt

  • 592 their individual characteristics (e.g., kind, cooperative). Receiving feedback that could be

  • 593 perceived as “rudeand “unkind” would be in direct contradiction to the cooperative ori-

  • 594 entation of an agreeable person. In other words, persons higher in agreeableness may be

  • 595 more sensitive to the damaging eVects of destructive interpersonal tactics and therefore

  • 596 express more anger in interactions that use these destructive tactics (Jensen-Campbell &

  • 597 Graziano, 2001; Suls et al., 1998).

  • 598 However, the inXuence of agreeableness on anger reactions held only for persons who

  • 599 were low on conscientiousness. Thus, individuals who have a prosocial, cooperative orienta-

  • 600 tion but are not able to “shift and focus their attention” may dwell on the negative interac-

  • 601 tion and its inappropriateness causing them to become angrier than their more conscientious

  • 602 peers. That being said, there was a trend (although not signiWcant) for persons lower on

  • 603 agreeableness to choose more aversive drink choices when angered. When persons were high

  • 604 on agreeableness, self-reported anger and prefrontal asymmetry were not associated with

  • 605 aversive drink choice. Thus, it may be possible that although agreeable persons may be more

  • 606 angered by rude comments, they are better able to control that anger to choose more socially

  • 607 acceptable responses. More research is needed to examine these interesting Wndings.

  • 608 4.1. Future directions

  • 609 The present study has a number of obvious strengths. First, the study used multiple

  • 610 methods to assess anger (i.e., prefrontal asymmetry and self-reported anger), which enabled

  • 611 us to examine the validity of our anger assessments. Our results partially replicated Har-

  • 612 mon-Jones and Sigelman (2001). For example, anger was related to mid-frontal but not lat-

  • 613 eral-frontal asymmetry in the negative feedback condition. Individuals in the negative

  • 614 feedback condition also showed greater lateral-frontal activity than did persons in the pos-

  • 615 itive feedback condition. Both self-reported anger and lateral-frontal activity were associ-

  • 616 ated with drink choice; mid-frontal activity, however, was not associated with drink choice

  • 617 in the negative feedback condition.

  • 618 Second, this study employed important controls and comparisons. First, this study

  • 619 examined prefrontal asymmetry while comparing it to parietal asymmetry. The anticipated

  • 620 diVerentiation between prefrontal asymmetry and parietal asymmetry was found. Second,

  • 621 parietal activity was not systematically related to self-reported anger or beverage choice in

  • 622 the negative feedback condition. Third, neither conscientiousness nor agreeableness was

  • 623 associated with parietal activity in the negative feedback condition (see Table 3). Fourth,

  • 624 conscientiousness moderated the prefrontal asymmetry-aggression link in the negative

ARTICLE IN PRESS

  • 18 L.A. Jensen-Campbell et al. / Journal of Research in Personality xxx (2006) xxx–xxx

  • 625 feedback condition; this moderating eVect was not replicated for parietal activity. Finally,

  • 626 the pattern of results only held for post-experimental (not baseline) measures in the nega-

  • 627 tive feedback condition.

  • 628 There are also several limitations in the current study as well. First, there was not

  • 629 enough power to detect higher-order interactions. For example, there were no signiWcant

  • 630 agreeableness £ anger interactions within the negative feedback condition. However, closer

  • 631 examination of the data suggests that the link between anger and aversive drink choice

  • 632 held only for persons with medium or low levels of agreeableness (see Table 4). A similar

  • 633 pattern was found for the moderating inXuence of conscientiousness on the mid-frontal

  • 634 asymmetry to aversive drink choice relationship.

  • 635 As stated previously, analyses were run separately by condition because of power issues

  • 636 associated with detecting three-way interactions (see Aiken & West, 1991, p. 139).

  • 637 Although having a comparison positive feedback condition was strength of this study,

  • 638 future research would require a much larger sample size to detect signiWcant three-way

  • 639 (and even some two-way) interactions. This study suggests that the moderating inXuences

  • 640 probably exist, but the small sample size limited our ability to Wnd them.

  • 641 Second, we did not Wnd a direct link between agreeableness and conscientiousness to

  • 642 aversive drink choice. Future research needs to assess the inXuences of conscientiousness

  • 643 and agreeableness on aggression using diVerent paradigms (e.g., paradigms used by Rich-

  • 645 between conscientiousness, agreeableness, and aggression (Gleason, Jensen-Campbell, &

  • 646 Richardson, 2004; Jensen-Campbell, 2006; Jensen-Campbell & Malcolm, 2006; Meier,

  • 647 Robinson, & Wilkowski, 2006; Workman, 2000). Although our operationalization of

  • 648 aggression is valid and has been used in previous research (e.g., Harmon-Jones & Sigelman,

  • 649 2001; Lieberman, Solomon, Greenberg, & McGregor, 1999), multiple converging assess-

  • 650 ments may help researchers gain a better insight into how personality may moderate the

  • 651 anger–aggression link.

  • 652 Finally, there was no direct measure of self-control in this study. Instead, this study

  • 653 focused on how conscientiousness and agreeableness moderate the anger–aggression link.

  • 654 Based on previous empirical data, it was theorized that the reason conscientiousness and

  • 655 agreeableness would moderate this known association is because these dimensions are pre-

  • 656 sumed to have temperamental origins in eVortful control (Ahadi & Rothbart, 1994). Future

  • 657 research, however, should not only focus on expanding the currently proposed moderator

  • 658 model, but should also examine how self-control processes associated with agreeableness and

  • 659 conscientiousness might mediate the link between Big Five traits and emotional and behav-

  • 660 ioral regulation. Some work in the area has already begun. For example, Meier et al. (2006)

  • 661 found that agreeableness was associated with the regulation of aggression-related primes

  • 662 (e.g., the activation of prosocial thoughts). In addition, D’Alfonso et al. (2000) found that

  • 663 manipulating the activation of the right prefrontal cortex through repetitive transcranial

  • 664 magnetic stimulation (rTMS) led participants to attentionally avoid angry faces. Finally,

  • 665 Harmon-Jones, Peterson, and Vaughn (2003) found that sympathy is associated with the

  • 666 reduction of relative left frontal asymmetry associated with anger. Finally, this study found

  • 667 that individuals higher on conscientiousness showed less left lateral-frontal activity while

  • 668 showing greater activation in the right lateral-frontal region when involved in a frustrating

  • 669 situation. It may be that high-conscientious persons are naturally better at shifting their

  • 670 attention away from frustrating situations and toward sympathic or prosocial thoughts,

  • 671 which in turn leads to avoidance of aggressive confrontations.

ARTICLE IN PRESS

L.A. Jensen-Campbell et al. / Journal of Research in Personality xxx (2006) xxx–xxx

  • 672 5. Conclusions

19