Relating momentary affect to the five factor model of ... russell suzuki.pdf · utility of the...

© 2003 Japanese Psychological Association. Published by Blackwell Publishing Ltd.

Japanese Psychological Research

2003, Volume 45, No. 2, 80–93

Munksgaard

ORIGINAL ARTICLE

Relating momentary affect to the five factor model of personality: A Japanese case

1

MICHELLE S. M.

YIK

2

Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong

JAMES A.

RUSSELL

Boston College, Chestnut Hill, Massachusetts, USA

NAOTO

SUZUKI

Department of Psychology, Faculty of Letters, Doshisha University, Karasuma Imadegawa, Kamigyo-ku, Kyoto 602-8580, Japan

Abstract:

This article shows that fundamental aspects of the structure of momentaryaffect are similar in Japanese and Canadian societies. We developed questionnaire scalesin Japanese in four different formats for assessing self-reported momentary affect. Scalescan be scored for dimensions defined by Feldman Barrett and Russell (1998), Thayer (1996),Larsen and Diener (1992), and Watson and Tellegen (1985). We administered these newlydeveloped affect scales and NEO five-factor inventory (Costa & McCrae, 1992) to a sampleof 450 Japanese respondents. The affect scales were found to be psychometrically soundand to be interrelated, as found with English-speaking Canadians. Dimensions could beintegrated into a two-dimensional affective space. Personality correlated with momentaryaffect, though not in the same pattern as found in Canada.

Key words:

Strucutre of Affect, Five Factor Model of Personality, Japanese.

For theoretical and practical reasons, psycho-logists are increasingly turning to the study ofaffect. Affect has been found to be an importantfactor in research laboratory, clinic, advertising,and workplace. There are three interrelatedpurposes in the present study. First, we test thegeneralizability to Japanese of the two-dimensionalaffective model developed in English. Second,we develop ready-to-use tools to assess momen-tary affect in Japanese. Finally, we conduct across-language comparison on the connectionsbetween momentary affect and personality.

A circumplex model of affect

In the past decade, various dimensional modelshave been proposed to describe the covari-ations of self-reported affective feelings inEnglish. Major models include Russell’s (1980)circumplex, Thayer’s (1996) energetic and tensearousal, Larsen and Diener’s (1992) eight com-binations of pleasantness and activation, andWatson and Tellegen’s (1985) Positive Affectand Negative Affect. These models were shownto fit comfortably within a two-dimensionalaffective space, characterized by two bipolar axes

1

Data reported in this article are part of Michelle Yik’s doctoral dissertation. We thank Yuka Mukai and Yuriko Takahasifor their help in translation.

2

Correspondence concerning this article should be sent to: Michelle Yik, Hong Kong University of Science andTechnology, Division of Social Science, Clear Water Bay, Kowloon, Hong Kong (Email: [email protected]).

JPR_036.fm Page 80 Wednesday, February 26, 2003 9:54 AM

© Japanese Psychological Association 2003.

Momentary affect in the Japanese

81

of Pleasant versus Unpleasant, and Activatedversus Deactivated (Feldman Barrett & Russell,1998; Yik, Russell, & Feldman Barrett, 1999).Figure 1 shows an empirical example of thatintegrated space in English (Russell, Yik, &Steiger, in press). On the right hand side arethe pleasant states; on the left hand side are theunpleasant states. On the upper half are theactivated states; on the lower half are the deac-tivated states. Within this two-dimensional space,any specific affective state is composed ofdifferent levels of Pleasantness and Activation.This space is a circumplex in which affectivedimensions fall in a circular ordering along theperimeter. The circumplical nature of affectivestates has received strong empiric support(Remington, Fabrigar, & Visser, 2000). Wedo not, however, assume that the structureof Figure 1 captures all of affect. Rather, wepropose it as a means of representing affect ata general level.

One question immediately arises: Can theintegrated space be generalized to other lan-guages, such as Japanese? Initial support forthe integrated structure was obtained fromprevious studies in Japanese. Ogawa, Takehara,Monchi, Fukui, and Suzuki (1999) found that

the two-dimensional space was useful to describethe ratings of facial expressions under ambigu-ous perceptual conditions (see also Ogawa &Suzuki, 2000). Affect scales were also developedto measure some vectors of the proposedintegrated space (Ogawa, Monchi, Kikuya, &Suzuki, 2000). In the present study, scalesdefining the affective dimensions in Figure 1were translated into Japanese. The resulting toolsare valuable for pursuing each of the originalmodels in Japanese. Together, they also allowedus to examine whether they can be integratedin the way indicated as they are in Figure 1.Integration relied on confirmatory factor ana-lysis, a powerful tool which estimates relationsamong variables relatively free of random andsystematic errors inherent in measurement.To take full advantage of confirmatory factoranalysis requires that each dimension be meas-ured in several different ways. We thereforedeveloped all scales in at least three differentresponse formats.

Predicting affect from personality

Our study also explored the link betweenmomentary affect (state) and personality (trait).Momentary affect obviously can be predicted

Figure 1. A circumplex model of affect. Fourteen unipolar affect constructs empirically placed in an integrated two-dimensional space via CIRCUM (Browne, 1992). Results are obtained from a study of 535 English-speaking Canadians. Adopted from Russell, Yik, and Steiger (in press).



82

M. S. M. Yik, J. A. Russell and N. Suzuki

by immediate context, but it is also obviousthat some people typically feel anxious, otherstypically feel happy, and still others typicallyfeel relaxed. In other words, one’s momentaryaffect can be predicted from one’s personality(e.g., Diener, 1984; Larsen & Ketelaar, 1991;McCrae & Costa, 1991). In the present study,we use the Japanese translation of NEO FiveFactor Inventory (NEO FFI; Costa & McCrae,1992) to measure personality.

Much work has linked Extraversion (E) andNeuroticism (N) to Watson and Tellegen’s (1985)Positive Affect and Negative Affect, respectively.The robustness of the findings led Tellegen(1985; Watson & Clark, 1984) to suggest thatE and N should be relabeled as “PositiveEmotionality” and “Negative Emotionality.”Fewer studies have been conducted to examinethe predictive utility of Agreeableness, Con-scientiousness, and Openness to Experience onaffect (Costa & McCrae, 1984; McCrae & Costa,1991; Watson & Clark, 1992). Positive relationswere reported between Openness to Experienceand positive affective states. Both Agreeable-ness and Conscientiousness were found to corre-late positively with positive affective states andnegatively with negative affective states.

The circumplex model of Figure 1 providesa simple but powerful way to summarizerelations between the affect variables and anyoutside variable, such as a personality variable.The principle is that any outside variable whichcorrelates reliably with one affect variable willcorrelate with the remaining affect variablesand the pattern of correlations will form a sinewave. That is, the magnitude of the correlationswill rise and fall in a sine pattern. Thus, it is theappearance of the sine wave, rather than thestatistical significance of individual correla-tions, which speaks directly to the integrity andutility of the circumplex model of affect.

Overview of the present study

Data from 450 Japanese respondents were usedto examine three issues: (a) The psychometricproperties of the translated affect scales; (b)the structure among those scales, specificallythe circumplex ordering; and (c) the relationof momentary affect to personality.

Method

Participants

Participants were 450 undergraduates (228men, 222 women) from Doshisha University.Their mean age was 19.69 (

SD

= 1.15). Parti-cipation was voluntary. Test administrationtook place during class time.

Procedure

Participants first completed an affect ques-tionnaire under the title “RememberedMoments Questionnaire” and then a person-ality questionnaire under the title “NEO FFI.”All questionnaires were in Japanese.

Affect questionnaire

Translation.

All instructions and scales weretranslated into Japanese by two bilinguals. Aback-translation procedure was adopted. First,one bilingual translated the English version ofthe affect questionnaire (Yik, Russell, & FeldmanBarrett, 1999) into Japanese. Second, anotherbilingual, who was blind to the English original,translated the Japanese version back intoEnglish. Translations were revised untilsatisfactory before we used them in the datacollection.

3

Instructions

. The front page of the batteryprovided general instructions under the title“Remembered Moments Questionnaire.”There were six versions of the questionnaire,each with a different anchoring time. The sixanchoring times were “before breakfast,”“after breakfast,” “before lunch,” “afterlunch,” “before dinner,” and “after dinner.”Participants were randomly assigned to one ofthe six instructions. For instance, the instruc-tions for one version were as follows:

3

The backtranslation procedure is not without prob-lems. The selection of equivalents is a matter of judg-ment. Different translators and different researchersroutinely produce slightly different translations. Inour case, we made our judgment call. We used dataanalysis as a means of rejecting items that failed tobehave properly.




83

… we need to ask you to remember aparticular moment. Please think back toyesterday. Specifically, recall the time just

before breakfast.

(If you didn’t have

breakfast

yesterday, simply recall that approximatetime of day.)

It is important that you remember a specificmoment accurately. So, please search yourmemory and try to recall where you were,what you were doing at that time, who youwere with, and what you were thinking.

Now select a particular moment that isespecially clear in your memory. (If youreally have no recollection of the time just

before breakfast

, please search your memoryfor the closest time that you do recallaccurately.)

In the other five versions, italicized wordswere replaced. The instructions then empha-sized that all subsequent questionnaires wereto be answered with respect to that selectedmoment of the day before.

Formats

. Participants then received a batteryof four questionnaires, each in a different for-mat, in the following order: (a) Semantic dif-ferential scales (SEM); (b) adjective format(ADJ), which was an adjective list accompaniedby a five-point Likert scale ranging from 1 “notat all” to 5 “extremely”; (c) “Agree-Disagree”format (AGREE), a list of statements with whichparticipants were asked to indicate their degreeof agreement, ranging from 1 “strongly dis-agree” to 5 “strongly agree”; and (d) “DescribesMe” format (DESCRIBE), a list of statements,for each of which participants were asked toindicate how well it described their feelings,ranging from 1 “not at all” to 4 “very well.”

The SEM format consisted of bipolarmeasures of Pleasure and Arousal translateddirectly from Mehrabian and Russell (1974).The remaining three questionnaires wereunipolar in format and each questionnaireincluded translated items from (a) FeldmanBarrett and Russell’s (1998) Current MoodQuestionnaire (CMQ) assessing Pleasant,Unpleasant, Activated, and Deactivated affect;

(b) Larsen and Diener’s (1992) ActivatedUnpleasant, Unactivated Unpleasant, ActivatedPleasant, and Unactivated Pleasant affect; (c)Thayer’s (1996) Energy, Tiredness, Tension,and Calmness; and (d) Watson, Clark andTellegen’s (1988) Positive Affect and NegativeAffect. Therefore, scores on these various scalescould be derived from our sample. The Japaneseversion of all affect measures is available fromthe first author upon request.

Personality measure

Our measure of personality was a 60-itemJapanese translation of NEO Five FactorInventory (NEO FFI; Costa & McCrae, 1992)developed by Shimonaka, Nakazato, Gondo,and Takayama (1999). NEO FFI is part of theirJapanese translation of the 240-item NEOPersonality Inventory Revised (NEO PI-R;Costa & McCrae, 1992) and is intended to offeran abbreviated measure for revised NEOpersonality inventroy (NEO PI-R). Responsesare made on a 5-point rating scale ranging from“Strongly Disagree” through to “Neutral” to“Strongly Agree.” Each of the five NEO FFIscales consisted of 12 items.

Cronbach’s alphas for the five personalityscales ranged from 0.63 to 0.85, indicating thatthey are internally consistent. The five scalesshowed the expected small, but reliable, corre-lations with each other. These psychometricproperties resemble those found in the admin-istration of NEO FFI in a similar Japanesesample (Shimonaka, et al., 1999) and with theoriginal English version (Costa & McCrae,1995).

Results and Discussion

The results are presented in four sections.First, we examine psychometric properties ofa priori scales. Second, we revise the scales tomore closely approximate the desired struc-ture, and report psychometric properties of therevised scales. Third, we examine the integratedstructure among the affect constructs originatingfrom different structural models. Fourth, weexamine the relation between affect (basedon the revised scales) and personality.



84


Individual measurement models

In this section, we examine the ability of thevarious a priori affect scales to assess theoriginal four structures from which the scaleswere developed. To examine how well each ofthe four structural models fit the Japanesedata, we used a confirmatory factor analysis,with each latent construct indicated by threescales with different response formats. For allfactor models, we estimated: (a) factor loadingbetween each manifest variable and its intendedlatent construct; (b) error term associated witheach manifest variable; (c) correlation betweenerror terms with the same response format;and (d) correlations between latent constructs.

Table 1 gives indices of fit for all the models.Models hypothesized by Feldman Barrett andRussell (1998), Thayer (1996), and Larsen andDiener (1992) all fit the data well, significantlybetter than did a comparison model in whichthe correlations among latent constructs were

fixed to 0.00. In contrast, the hypothesizedmodel for Watson and Tellegen (1985) fit thedata poorly. Yik et al. (1999) found that bipolarversions of Watson and Tellegen’s dimensions(Positive Affect and Negative Affect) produceda reasonable measurement model. To replicatethe findings of Yik et al., Larsen and Diener’sUnactivated Unpleasant was used as thebipolar opposite of Positive Affect, UnactivatedPleasant the opposite of Negative Affect. Toassess the bipolar version of Watson andTellegen’s model, we specified a confirmatoryfactor analysis with four latent constructs,each indicated by three scales with differentresponse formats. The hypothesized model fitthe data modestly:

χ

2

(30,

N

= 450) = 318.55,root mean square error of approximation(RMSEA) = 0.15, adjusted goodness of fit(AGFI) = 0.73, comparative fix index (CFI) =0.94. This model also fit the data significantlybetter than the comparison model:

∆χ

2

(6,

N

=

Table 1. Indices of fit for a priori and revised measurement models (N = 450)

Model χ2 df RMSEA AGFI CFI

CMQ – a prioriModel with Correlated Constructs 194.62 30 0.11 0.83 0.96Model with Correlations between

Constructs fixed to zero829.02 36 0.20 0.56 0.80

CMQ – RevisedModel with Correlated Constructs 173.40 30 0.10 0.85 0.95Model with Correlations between


Thayer’s Unipolar Constructs – a prioriModel with Correlated Constructs 185.93 30 0.11 0.84 0.96Model with Correlations between


Larsen & Diener’s Unipolar Constructs – a prioriModel with Correlated Constructs 170.45 30 0.10 0.85 0.97Model with Correlations between


Watson & Tellegen’s Unipolar Constructs – a prioriModel with Correlated Constructs 161.13 5 0.25 0.58 0.93Model with Correlations between


Constructs fixed to zero.CMQ, Current Mood Questionnaire; RMSEA, root mean square error of approximation; AGFI, adjusted goodness of fit; CFI, comparative fit index.




85

450) = 514.89,

p

< 0.001, and RMSEA changedfrom 0.15 to 0.22.

To sum up, the structures proposed by eachof the original authors based on English datawas supported in the present Japanese data, withsome difficulties encountered for Watson andTellegen’s (1985) variables. Individual constructs

were adequately measured by the three scaleswith different response formats. Constructs wererelated to each other approximately as predicted.Since the Feldman Barrett and Russell (1998)model forms the core of the proposed integratedspace, the model estimates are given in Table 2.Parameter estimates for other structural models

Table 2. Measurement models for a priori (and revised) Current Mood Questionnaire (CMQ) scales: Confirmatory factor analyses (N = 450)

Construct Format Pleasant Unpleasant Activated Deactivated M SD

Standardized Factor LoadingPleasant ADJ 0.90* 2.76 1.16

(.87*) (2.83) (1.23)Pleasant AGREE 0.97* 3.21 1.12

(.96*) (3.27) (1.13)Pleasant DESCRIBE 0.75* 2.41 0.82

(.82*) (2.58) (1.01)Unpleasant ADJ 0.93*

(.90*) 1.78(1.85)

0.94(1.05)

Unpleasant AGREE 0.90* (.89*)

1.93(1.93)

1.08(1.08)

Unpleasant DESCRIBE 0.90* (.88*)

1.85(1.85)

0.89(.89)

Activated ADJ 0.78* (.58*)

2.18(2.17)

0.97(1.03)

Activated AGREE 0.86* (.58*)

2.36(2.55)

0.89(1.09)

Activated DESCRIBE 0.61* (.41*)

1.78(1.67)

0.63(.87)

Deactivated ADJ 0.71* (.65*)

2.54(2.74)

0.91(1.01)

Deactivated AGREE 0.62* (.56*)

2.64(2.64)

0.87(.87)

Deactivated DESCRIBE 0.85* (.91*)

2.31(2.23)

0.66(.77)

Interfactor CorrelationPleasant –

(–)Unpleasant –.78* –

(–.81*) (–)Activated –.31* 0.63* –

(.18*) (.08) (–)Deactivated 0.00 –.10 –.49* –

(–.11*) (.06) (–.72*) (–)

Figures without parentheses are estimates for a priori CMQ scales; figure in parentheses are estimates for revised CMQ scales. Possible mean scores range from 1 to 5 for Adjective and “Agree-Disagree” formats; 1–4 for “Describes Me” format.p ≤ 0.01.



86


are available from the first author uponrequest.

Revising the affect scales

Because the scales defining the 14 affectconstructs are borrowed directly from Englishby translation, it is possible that these a prioriscales were not adequate indicators of theintended underlying latent constructs and mightlack adequate internal consistency. Suchpsychometric problems could obscure thestructural appearance of the models tested andmay worsen the fit indices and thus the meas-urement models. In order to define the latentconstructs and thus the structural modelstested by culturally appropriate scales, effortsare therefore directed to revising the scales onthe basis of measurement models and reliabilityanalyses.

Revising the Current Mood Questionnaire scales

.Pleasant, Unpleasant, Activated, and Deact-ivated – these four affect constructs are thecornerstones of the two-dimensional spaceproposed in the present study. Thus, it is ofparamount importance to ensure that thepattern of correlations among these four alignswith that expected. In order to ensure the properplacements of the four constructs withinthe two-dimensional space, the Pleasant andUnpleasant scales must be relatively independ-ent of the Activation dimension (neitheractivated nor deactivated) and that Activatedand Deactivated scales must be relativelyindependent of the Pleasantness dimension(neither pleasant nor unpleasant). We thusrevised these 12 scales (3 scales

×

4 constructs)with two interrelated purposes in mind: (a) thenegative correlations between the hypothesizedbipolar opposites (Pleasant

vs.

Unpleasant,Activated

vs.

Deactivated) should be maximized;and (b) the correlations between non-bipolarpairs (such as Pleasant

vs.

Activated) shouldbe minimized.

Any revision procedure can be accused ofcapitalizing on chance. We therefore took stepsto minimize this possibility: (a) No items wereallowed to switch from one scale to another;and (b) items could only be dropped (but not

added) in the revision procedure. With thesecriteria in mind, we found that revisions wereneeded. Of 12 scales (4 constructs

×

3 responseformats), 10 were revised.

As shown in Table 1, the revised CMQscales fit the hypothesized model better thanhad the a priori scales. Further, the revisedCMQ model fit the data significantly betterthan did a comparison model with correlationsamong latent constructs fixed to 0.00:

∆χ

2

(6,

N

= 450) = 536.33,

p

< 0.001, and RMSEA changedfrom 0.10 to 0.19. Parameter estimates aregiven in parentheses in Table 2 and the revisedscales were used in subsequent analyses.

Revising other affect scales

.

To maintain thesimilarity between the Japanese and Englishversions of the scales developed by otherauthors, we used a more conservative pro-cedure in the present revision. Reliability esti-mates for scales defining Thayer’s (1996),Larsen and Diener’s (1992), and Watson andTellegen’s (1985) constructs were examined.The purpose was to make sure that the scaleswere internally consistent (with a minimumalpha of 0.60) and that they were a reasonableindicator of the intended constructs (with aminimum factor loading of 0.70). All but oneof Thayer’s scale passed the reliability criterion;one item was dropped from that scale whichwas used in subsequent analyses. All scalespassed the 0.70 loading criterion.

The full two-dimensional affective space

A correlation matrix including all the affectscales examined here showed substantialcorrelations across different structural models.Our hypothesis was that all structures aremappable onto one common integrated space.In the following, we adopted two ways to testthis convergence across structures.

Structural equation models

. One way to demon-strate the convergence of constructs ofdifferent origins was to use Pleasant versusUnpleasant and Activated versus Deactivatedaxes as exogenous variables to predict allother affect constructs. By treating all otherconstructs as endogenous variables, we could




87

test the hypothesis that the two axes explainmost of the reliable variance in other affectconstructs.

First, we specified a confirmatory factor model(what we call Model 1) with two latent con-structs, corresponding to the bipolar axes ofPleasant versus Unpleasant and Activatedversus Deactivated affect. Each latent constructwas indicated by the bipolar versions of itsthree scales with different response formats.The semantic differential scale of Pleasurewas specified to load on the Pleasant versusUnpleasant construct; the semantic differentialscale of Arousal was to load on the Activatedversus Deactivated construct. The latent corre-lation between the two axes was fixed to 0.00.Model 1 fit the data moderately well:

χ

2

(16,

N

= 450) = 111.86, RMSEA = 0.12, AGFI = 0.86,CFI = 0.97. Model estimates were used in sub-sequent analyses.

In the following structural equation models,parameter estimates on the exogenous sidewere adopted from Model 1 in the precedingparagraph. In each analysis, we estimated(a) loading between a manifest variable andthe endogenous construct; (b) regression weights

of the endogenous construct on the exogenousconstructs; and (c) percentage of variance ex-plained by the exogenous constructs for eachendogenous construct. We conducted a separateanalysis for each of the six bipolar constructs(2 from Thayer, 2 from Larsen and Diener, and2 from Watson and Tellegen). Results aresummarized in Table 3.

All affect constructs were substantiallyexplained by the two axes. The mean varianceexplained was 92% for the pleasant activatedaffects and 91% for the unpleasant activatedaffects. The pattern of relations between theexogenous variables and the endogenousvariables was approximately as expected inFigure 1. Consistent with results obtained inEnglish, the four structures can be comfortablyintegrated into a two-dimensional space.

CIRCUM

.

Another way to demonstrate theconvergence of the constructs across differentmodels was to portray the full representa-tion of all constructs simultaneously within atwo-dimensional space. To do so, we usedCIRCUM, a structural equation modelingprogram developed by Browne (1992) to

Table 3. Affect Constructs Explained by the Revised Pleasant versus Unpleasant and Activated versus Deactivated Axes (N = 450)

Construct

Regression Weight

Variance explained (SE)

Indices of Fit

Pleasant vs.Unpleasant

Activated vs.Deactivated χ2 RMSEA AGFI CFI

Pleasant Activated vs. Unpleasant Deactivated affectEnergy vs. Tiredness† 0.64 0.69 89 (1.6) 299.84 0.10 0.87 0.95Activated Pleasant vs.Unactivated Unpleasant‡

0.72 0.65 94 (1.3) 241.37 0.09 0.90 0.96

High vs. Low Positive Affect§ 0.61 0.74 93 (1.5) 296.65 0.10 0.87 0.95

Unpleasant Activated vs. Pleasant Deactivated affectTension vs. Calmness† –.74 0.58 88 (1.7) 326.27 0.11 0.86 0.95Activated Unpleasant vs. Unactivated Pleasant‡

–.87 0.41 92 (1.2) 297.33 0.10 0.87 0.96

High vs. Low Negative Affect§ –.87 0.40 92 (1.2) 283.44 0.09 0.88 0.96

All regression coefficients are significant at .001 level, which is equivalent to an overall alpha level of less than .01 (by Bonferroni procedure, 12 regression weights × 0.001 = 0.012).†Thayer (1996); ‡Larsen and Diener (1992); §Watson and Tellegen (1985).RMSEA, root mean square error of approximation; AGFI, adjusted goodness of fit; CFI, comparative fit index.

JPR_036.fm 7 Wednesday, February 26, 2003 9:54 AM


88


examine how well our data conformed to acircumplex structure. This program providesfit indices and angular position for each affectvariable.

This analysis uses the unipolar affect scales.(The semantic differential scales were not usedin this analysis.) First, a score was created foreach of the 14 unipolar constructs by summingthe

z

-scores of its three scales with differ-ent response formats. A 14

×

14 correlationmatrix was then computed with the resultingsums and was submitted to the maximum like-lihood estimation using CIRCUM. Pleasantwas designated as the reference variable(its location was fixed at 0

°

). The locations ofother variables were then estimated relativeto Pleasant. The communality estimates of allvariables were left free to vary. No constraintswere put on the minimum common scorecorrelation.

The analysis converged on a solution in 12iterations. Three free parameters were specifiedin the correlation function equation; additionalfree parameters did not improve the modelfit. The final model had a total of 44 freeparameters and 61 degrees of freedom. Thedata fit the model only moderately well:

χ

2

(61,

N

= 450) = 355.45, RMSEA = 0.10.The results are shown in Figure 2. The four

cornerstone variables (Pleasant, Unpleasant,Activated, and Deactivated) were located closeto the predicted values: With Pleasant fixed at0

°

, Activated was 76

°

away, Unpleasant was167

°

away, and Deactivated was 262

°

away.Hypothesized bipolar opposites were locatedclose to the predicted values: Pleasant was 167

°

from its bipolar opposite, Unpleasant. Activatedwas 186

°

from its bipolar opposite, Deactivated.Constructs developed by various authors fellremarkably close to what we see in Figure 1.

Figure 2. Japanese scales of momentary affect. A circumplex representation of 14 unipolar affect constructs viaCIRCUM (Browne, 1992). Communality was left free to vary. Figures given are estimates of polar angleswith the 95% confidence intervals in parentheses.




89

Summary

.

We set out to test the hypothesisof convergence of affect constructs acrossdifferent structural models by structuralequation models and the CIRCUM analyses.Results from both analyses showed that a two-dimensional space defined by Pleasant versusUnpleasant and Activated versus Deactivatedaxes was able to integrate affect constructsoriginating from different authors. Further,variables fell at various angles within thespace. The results are consistent with a cir-cumplex, which predicts variables to fall at anyplace along a circle.

Relating affect to personality

In this section, we examine the connectionsbetween affect and personality. First, weexamine the ability of personality in predictingmomentary affect in a series of structuralequation models. Second, we place eachpersonality dimension into the integrated affec-tive space.

Predictive utility of personality

.

To investigatethe ability of personality to predict affect, wetreated personality as exogenous and affectas endogenous. All exogenous variables weremanifest variables. Each of the eight bipolaraffect variables served, in turn, as an endogenousvariable. In all, we computed 16 structuralequation models. For all structural equationmodels, we estimated: (a) loading between amanifest variable and the endogenous con-struct; (b) error term associated with eachmanifest variable; (c) regression weights ofthe exogenous variables on the endogenousvariables; and (d) a percentage of varianceexplained by the exogenous variables for eachendogenous construct.

The question was, for each bipolar affectvariable, which combination of personalitydimensions was the best predictor. In Table 4,we compared two: the E-and-N Model (Eand N) versus the FFM (FFM; N, E, O, A,and C). The E-and-N model fit the data well.The mean RMSEA was 0.03. The varianceexplained ranged from 2.7% to 6.7%, with amean of 5.0%. The FFM fit the data even bet-ter. The mean RMSEA was 0.02. The variance

explained ranged from 4.1% to 8.0%, with amean of 6.5%. Because the E-and-N modelwas nested with the FFM, we report the

χ

2

dif-ference between the two. In all eight cases, the

χ

2

change statistic favored the FFM. Only onecase showed a significant difference indicatinga reliable improvement in model fit by addingO, A, and C to E and N. Consistent withfindings in English, the five personality factorsyielded higher predictive ability than did the Eand N alone.

Structural convergence of affect and personality

.The circumplex provides a powerful predictionabout the pattern of correlations between theset of affect variables and any outside variable.The correlations between any particular per-sonality dimension and all 14 affect constructsshould fit a sine function. A sine wave can thenbe used to evaluate the integrity of the two-dimensional space and to locate the personalitydimensions within that space. We thereforefit a sine function to each set of correlationrelating a personality dimension with all 14affect constructs.

We first computed a score for each of the 14unipolar affect constructs by summing the z-scores of its three constituent scales. Second,we computed correlations between each FFMdimension and the affect constructs. In Figure3, the affect dimensions are represented on theabscissa by their angles within the circumplexderived from the CIRCUM analyses describedin the preceding section. A sine function fit thedata well for all personality dimensions.

Browne’s (1992) CIRCUM procedureprovides a maximum likelihood estimate of wherewithin the two-dimensional affective spaceeach personality dimension falls. For eachpersonality dimension, three figures are given.The angle estimates the location within thecircumplex for the personality variable. Thezeta estimates the correlation between the per-sonality dimension and the affect vector at theangle specified. The VAF examines the modelfit. Results are shown in Table 5. Comparablevalues from an English sample are also shown.Personality and momentary affect were linkedin different ways between Japanese and English.



90


Table 4. Predicting affect from personality: A comparison of Japanese with English (N = 450)

Table 5. Empirical location of personality dimensions in the two-dimensional affective space via CIRCUM-extension

Regression weight Variance explained by E-and-N’s model (SE)

Variance Explained

by FFM (SE)N E O A C

Two AxesPleasant vs. Unpleasant† –.15 0.01 0.02 0.17* 0.04 4.9 (2.0) 7.5 (2.5)

–.32* 0.04 0.06 0.06 0.05 13.0 (2.8) 13.9 (2.8)Activated vs. Deactivated† 0.01 0.20* 0.05 –.03 0.08 4.4 (2.2) 5.2 (2.3)

0.08 0.18* 0.05 –.03 0.06 3.0 (1.5) 3.6 (1.6)Pleasant Activated vs. Unpleasant Deactivated affectsEnergy vs. Tiredness‡ –.12 0.14 0.05 0.05 0.11 6.5 (2.3) 8.0 (2.5)

–.17* 0.04 0.06 –.01 0.07 4.0 (1.7) 4.7 (1.8)Activated Pleasant vs. Unactivated Unpleasant§ –.13 0.14 0.07 0.05 0.08 6.7 (2.3) 8.0 (2.5)

–.20* 0.07 0.10 0.01 0.07 6.5 (2.1) 8.0 (2.3)High vs. Low Positive Affect¶ –.13 0.13 0.08 0.04 0.08 5.6 (2.2) 6.9 (2.4)

–.21 0.06 0.13 0.01 0.12 7.1 (2.2) 9.8 (2.5)

Unpleasant Activated vs. Pleasant Deactivated affect

Tension vs. Calmness‡ 0.15 0.08 –.03 –.12 –.03 2.7 (1.6) 4.1 (1.9)0.34* 0.08 –.02 –.04 0.04 10.1 (2.5) 10.5 (2.6)

Activated Unpleasant vs. Unactivated Pleasant§ 0.19* 0.04 –.01 –.14 0.01 4.5 (2.0) 6.1 (2.2)0.38* 0.07 –.05 –.04 0.00 13.4 (2.8) 13.8 (2.8)

High vs. Low Negative Affect¶ 0.19* 0.04 –.03 –.14 –.02 4.5 (2.0) 6.3 (2.3)

0.38* 0.07 –.04 –.06 –.01 13.8 (2.8) 14.3 (2.8)

Figures in italic are results in an English sample of N = 535 (Yik, 1998). Figures in parentheses are the standard errors. N, Neuroticism; E, Extraversion; O, Openness to Experience; A, Agreeableness; C, Conscientiousness. Significance level was set at .001 in order to achieve an overall alpha of less than .05 (by Bonferroni procedure, 40 regression weights × 0.001 = 0.04).†Feldman Barrett and Russell (1998); ‡Thayer (1996); §Larsen and Diener (1992); ¶Watson and Tellegen (1985).p ≤ 0.001.

Japanese sample (N = 450) English sample (N = 535)†

Angle Zeta VAF Angle Zeta VAF

Neuroticism 176° 0.31 97 176° 0.47 98Extraversion 24° 0.21 84 28° 0.22 84Openness to Experience 63° 0.07 –‡ 32° 0.10 86Agreeableness 352° 0.24 87 355° 0.13 75Conscientiousness 15° 0.18 90 18° 0.17 84

Angle refers to the estimated angular position of the personality dimension within the two-dimensional affective space. Zeta refers to the estimated communality index for the personality dimension and indicates the correlation between the personality dimension and the common score. Model fit for placing a personality dimension within the circumplex was assessed by the Variance Accounted for (VAF).†Russell, Yik, and Steiger (in press).‡Value cannot be computed in CIRCUM-extension.



Momentary affect in the Japanese 91

The angles obtained were not identical betweenthe two samples.

Meyer and Shack (1989) suggested thatresearchers could use personality dimensionsto locate the fundamental dimensions of affect.Their prediction was that personality, or at leastE and N, would fall at 45° and 135° in the spaceof Figure 2. Their prediction was borne out neitherin Japanese nor in English. Deviations fromthese predictions were large, and the horizontal

axis is at least as close. From our perspective,it is the entire structure rather than on specificdimensions within it which is fundamental.

Conclusion

The present findings lend support to the viabil-ity of the structural model in Figures 1 and 2as an integration of various dimensional modelsfor momentary affect in both English andJapanese. To compare the Japanese results(Figure 2) with those in English (Figure 1), onecan simply superimpose the two figures on topof each other. It becomes immediately obvioushow similar the empiric placements of theaffect variables are. Indeed, the rank order ofthe 14 variables is almost identical in Englishand Japanese (rank order correlation = 0.98).Further, the locations of the variables agreevery well with the original authors’ conceptu-alizations: For instance, Thayer (1996) definedhis Energy scale as pleasant activation and itindeed fell in the pleasant activated quadrantin Figure 2.

The present study adopted the “imposed-etic”approach (Berry, 1969) in which translationsof scales originating from English were admin-istered to a sample of Japanese respondents.This approach emphasizes similarities acrosscultures and can be blind to indigenous con-structs or processes. Given the richness of theemotion lexicon of Japanese, the possibilityremains that additional affect dimensions wouldemerge with more indigenous items. Resultsobtained in the present study represent a firststep towards studying affect and its externalcorrelates in the Japanese culture. Here affectwas studied at a broad general level high in theaffect hierarchy, and further studies are muchneeded to examine more specific affectivedimensions at a lower level in a hierarchy. Wesuspect that cultural differences will be obviousthe lower one goes in that hierarchy.

In addition to studying the structure of affectand its relation to personality, there are a hugenumber of questions about affect. One of ourgoals in the present study was therefore toprovide ready-to-use, psychometrically soundaffect measures for use in Japanese. We end

Figure 3. The correlation of affect variables with apersonality dimension as a function of theangle within the circumplex for 14 affectvariables. The value for the affect variableat 0° is repeated at 360° to show the com-plete sine wave. N, Neuroticism; E, Extra-version; O, Openness to Experience; A,Agreeableness; C, Conscientiousness.



92 M. S. M. Yik, J. A. Russell and N. Suzuki

here by noting that these scales provide a briefand efficient means of capturing affect foundfor Japanese-speaking respondents. Complet-ing all 44 scales takes about 25 min. In basicresearch on structural relations between affectand other psychological variables in whichmeasurement errors must be minimized, werecommend all 44 scales. In many researchproblems, it would be more practical to use oneof the three response formats. If even morebrevity is required, the four scales for Pleasant,Unpleasant, Activated, and Deactivated affectsuffice to predict scores on the remainingdimensions. So, these scales can be used for agreat variety of purposes with minimal cost.

References

Berry, J. W. (1969). On cross-cultural comparability.International Journal of Psychology, 4, 119–128.

Browne, M. W. (1992). Circumplex models forcorrelation matrices. Psychometrika, 57, 469–497.

Costa, P. T. Jr, & McCrae, R. R. (1984). Personalityas a lifelong determinant of well-being. InC. Malastesta & C. Izard (Eds.), Affective pro-cesses in adult development and aging (pp. 141–157). Beverly Hills, CA: Sage.

Costa, P. T., Jr, & McCrae, R. R. (1992). The RevisedNEO Personality Inventory (NEO-PI-R) andNEO Five-Factor Inventory (NEO-FFI) pro-fessional manual. Odessa, FL: PsychologicalAssessment Resources.

Costa, P. T., Jr, & McCrae, R. R. (1995). Solid groundin the wetlands of personality: A reply to Block.Psychological Bulletin, 117, 216–220.

Diener, E. (1984). Subjective well-being. Psycholo-gical Bulletin, 95, 542–575.

Feldman Barrett, L., & Russell, J. A. (1998). Inde-pendence and bipolarity in the structure ofcurrent affect. Journal of Personality and SocialPsychology, 74, 967–984.

Larsen, R. J., & Diener, E. (1992). Promises andproblems with the circumplex model of emotion.In M. S. Clark (Ed.), Review of Personality andSocial Psychology: Emotion (Vol. 13, pp. 25–59). Newbury Park, CA: Sage.

Larsen, R., & Ketelaar, T. (1991). Personality andsusceptibility to positive and negative affect.Journal of Personality and Social Psychology,61, 132–140.

McCrae, R. R., & Costa, P. T. Jr (1991). AddingLiebe und Arbeit: The full five-factor model

and well-being. Personality and Social Psycho-logy Bulletin, 17, 227–232.

Mehrabian, A., & Russell, J. A. (1974). An approachto environmental psychology. Cambridge, MA:MIT Press.

Meyer, G. J., & Shack, J. R. (1989). Structural con-vergence of mood and personality: Evidence forold and new directions. Journal of Personalityand Social Psychology, 57, 691–706.

Ogawa, T., Monchi, R., Kikuya, M., & Suzuki, N.(2000). Development of the general affect scales.Japanese Journal of Psychology, 71, 241–246.

Ogawa, T., & Suzuki, N. (2000). Emotion space asa predictor of binocular rivalry. Perceptual andMotor Skills, 90, 291–298.

Ogawa, T., Takehara, T., Monchi, R., Fukui, Y., &Suzuki, N. (1999). Emotion space under condi-tions of perceptual ambiguity. Perceptual andMotor Skills, 88, 1379–1383.

Remington, N. A., Fabrigar, L. R., & Visser, P. S.(2000). Reexamining the circumplex model ofaffect. Journal of Personality and Social Psychology,79, 286–300.

Russell, J. A. (1980). A circumplex model of affect.Journal of Personality and Social Psychology,39, 1161–1178.

Russell, J. A., Yik, M. S. M., & Steiger, J. H. (inpress). A 12-point circumplex model of affect.Manuscript, Submitted for Publication.

Shimonaka, Y., Nakazato, K., Gondo, Y., &Takayama, M. (1999). Revised NEO PersonalityInventory (NEO-PI-R) and NEO Five-FactorInventory (NEO-FFI) manual for the Japaneseversion Tokyo: Tokyo Shinri. (In Japanese.)

Tellegen, A. (1985). Structures of mood and person-ality and their relevance to assessing anxiety,with an emphasis on self-report. In A. H. Tuma,& J. D. Maser (Eds.), Anxiety and anxiety dis-orders (pp. 681–706). Hillsdale, NJ: Erlbaum.

Thayer, R. E. (1996). The origin of everyday moods:Managing energy, tension, and stress. NewYork: Oxford University Press.

Watson, D., & Clark, L. A. (1984). Negative affect-ivity: The disposition to experience aversiveemotional states. Psychological Bulletin, 96,465–490.

Watson, D., & Clark, L. A. (1992). On traits andtemperament: General and specific factors ofemotional experience and their relation to theFive-Factor Model. Journal of Personality, 60,441–476.

Watson, D., Clark, L. A., & Tellegen, A. (1988).Development and validation of brief measuresof positive and negative affect: The PANASscales. Journal of Personality and Social Psycho-logy, 54, 1063–1070.



Momentary affect in the Japanese 93

Watson, D., & Tellegen, A. (1985). Toward aconsensual structure of mood. PsychologicalBulletin, 98, 219–235.

Yik, M. S. M. (1998). A circumplex model of affectand its relation to personality: A five-language study.Unpublished doctoral Dissertation, Vancouver,British Columbia: University of British Columbia.

Yik, M. S. M., Russell, J. A., & Feldman Barrett, L.(1999). Structure of current affect: Integrationand beyond. Journal of Personality and SocialPsychology, 77, 600–619.

(Received November 29, 2000; accepted May 17, 2002)


Relating momentary affect to the five factor model of ... russell suzuki.pdf · utility of the...

Documents

Transcript of Relating momentary affect to the five factor model of ... russell suzuki.pdf · utility of the...