30
KAREN J. CROOKER and JANET P. NEAR HAPPINESS AND SATISFACTION: MEASURES OF AFFECT AND COGNITION? (Accepted 11 November 1997) ABSTRACT. Analyses of data from six national samples of adult respondents indi- cated that happiness could be predicted better from cognitive measures of domain satisfaction and work attitudes than from a measure of positive affect, thereby calling into question the widely accepted argument that satisfaction measures are cognitive and happiness measures affective in orientation. Perhaps distinctions between cognitive and affective measures are illusory in studies of subjective well being. Studies of subjective well being have typically focused on life sat- isfaction to the exclusion of happiness, or added the two together in some sort of combined scale (e.g., Diener, 1984, 1994; McKennell, Atkinson and Andrews, 1980). One problem with this is that happi- ness and life satisfaction may in fact measure different constructs. While they are typically correlated at about 0.57 (McKennell et al., 1980) and previous research tends to treat them as though they were interchangeable (e.g., Veenhoven, 1984) or at least closely related (e.g., Wheeler, 1991), there has been some speculation that they dif- fer in one important regard. Beginning with Campbell (1976), it has been argued that happiness is an affective construct, while satisfac- tion is a cognitive construct, with a more evaluative tone (Andrews and McKennell, 1980; Brief and Roberson, 1989; McKennell, 1978; McKennell and Andrews, 1980; Organ and Near, 1985). Thus, two potentially different constructs may be involved. This is important because McKennell’s work suggested that affective measures generally contribute little additional variance Preliminary analysis of one of six data sets presented here appeared earlier in the Proceedings of the Fifth Quality-of-Life Marketing Conference, Williamsburg, VA, 1995. The data in this article were made available by the ISR Social Science Archive and the Inter-university Consortium for Political and Social Research. The 1972 data was collected by F.M. Andrews and S.B. Withey of the Survey Research Center, Institute for Social Research, The University of Michigan. The original collectors of the data, the Archive, and the Consortium bear no responsibility for the analysis or interpretation presented here. We wish to thank Dennis Organ and an anonymous reviewer for comments on an earlier draft of this paper. Social Indicators Research 44: 195–224, 1998. c 1998 Kluwer Academic Publishers. Printed in the Netherlands.

Happiness and Satisfaction: Measures of Affect and Cognition?

Embed Size (px)

Citation preview

Page 1: Happiness and Satisfaction: Measures of Affect and Cognition?

KAREN J. CROOKER and JANET P. NEAR

HAPPINESS AND SATISFACTION: MEASURES OF AFFECTAND COGNITION??

(Accepted 11 November 1997)

ABSTRACT. Analyses of data from six national samples of adult respondents indi-cated that happiness could be predicted better from cognitive measures of domainsatisfaction and work attitudes than from a measure of positive affect, therebycalling into question the widely accepted argument that satisfaction measures arecognitive and happiness measures affective in orientation. Perhaps distinctionsbetween cognitive and affective measures are illusory in studies of subjective wellbeing.

Studies of subjective well being have typically focused on life sat-isfaction to the exclusion of happiness, or added the two together insome sort of combined scale (e.g., Diener, 1984, 1994; McKennell,Atkinson and Andrews, 1980). One problem with this is that happi-ness and life satisfaction may in fact measure different constructs.While they are typically correlated at about 0.57 (McKennell et al.,1980) and previous research tends to treat them as though they wereinterchangeable (e.g., Veenhoven, 1984) or at least closely related(e.g., Wheeler, 1991), there has been some speculation that they dif-fer in one important regard. Beginning with Campbell (1976), it hasbeen argued that happiness is an affective construct, while satisfac-tion is a cognitive construct, with a more evaluative tone (Andrewsand McKennell, 1980; Brief and Roberson, 1989; McKennell, 1978;McKennell and Andrews, 1980; Organ and Near, 1985). Thus, twopotentially different constructs may be involved.

This is important because McKennell’s work suggested thataffective measures generally contribute little additional variance

? Preliminary analysis of one of six data sets presented here appeared earlier intheProceedings of the Fifth Quality-of-Life Marketing Conference, Williamsburg,VA, 1995. The data in this article were made available by the ISR Social ScienceArchive and the Inter-universityConsortium for Political and Social Research. The1972 data was collected by F.M. Andrews and S.B. Withey of the Survey ResearchCenter, Institute for Social Research, The University of Michigan. The originalcollectors of the data, the Archive, and the Consortium bear no responsibility forthe analysis or interpretation presented here. We wish to thank Dennis Organ andan anonymous reviewer for comments on an earlier draft of this paper.

Social Indicators Research44: 195–224, 1998.c 1998Kluwer Academic Publishers. Printed in the Netherlands.

Page 2: Happiness and Satisfaction: Measures of Affect and Cognition?

196 KAREN J. CROOKER AND JANET P. NEAR

to subjective well being beyond that contributed by the so calleddomain satisfactions, including job satisfaction, leisure satisfactionand so on (e.g., McKennell et al., 1980). The measure of subjectivewell being in this case was satisfaction, however, rather than happi-ness, so it was perhaps not surprising that affective variables werenot strongly related to a dependent variable that had been classifiedas cognitive rather than affective. A different approach would be totreat happiness as the dependent variable and then assess the varianceexplained by affect, when the effects of other cognitive independentvariables are controlled. If we find that the affect variables explainno incremental variance in happiness, beyond that explained by thecognitive variables, it may suggest problems in the classificationscheme that has identified happiness as affective in orientation andsatisfaction as cognitive.

This question is of more than methodological importance to ourunderstanding of the development of subjective well being over thecourse of careers and life spans. The relationships among variousdomain satisfactions, including job satisfaction, and overall life sat-isfaction, has been well documented (see Rain, Lane and Steiner,1991; Rice, Near and Hunt, 1980); further, these relationships canbe shown to change over the course of work careers and life stages.It is less clear what role measures of affect, including happiness,play in this process, although different theories have been put for-ward (Brief, Butcher and Robinson,1995; Clark and Watson, 1991;Judge and Eretz, 1993). For example, it has been well documentedthat satisfaction increases with age (although not always in a linearfashion) but happiness seems to decline with age. Assuming thatthe two measure the same construct, then, would obviously result infaulty interpretation of the phenomenon (Organ and Near, 1985). Onthe other hand, assuming that satisfaction is a cognitive measure andhappiness an affective measure, without verification, overly simpli-fies an issue that may be far more complex. It is important to assess,then, relationships among variables that purportedly represent bothaffective and cognitive orientations.

Research Question

Our prediction is that affect measures will explain significant incre-mental variance in happiness, beyond that explained by cognitive

Page 3: Happiness and Satisfaction: Measures of Affect and Cognition?

HAPPINESS AND SATISFACTION 197

TABLE IDemographics for the samples

A&W General Social Surveys

Dimension 1972 1973 1975 1989 1990 1991

% male 43.40 46.60 45.00 42.90 44.00 41.90% white 88.60 87.00 88.80 85.80 83.80 83.30average age/years 44.10 44.10 44.31 45.44 45.96 45.63% married 67.00 71.50 67.20 55.10 53.00 53.00average # children 2.20 2.11 1.95 1.87 1.90% currently employed 55.00 52.80 52.20 60.30 63.00 59.10average # years education 11.60 11.68 12.73 12.82 12.88N 1072 1504 1490 1537 1372 1517

variables. This study extends earlier work by McKennell (1978) byusing happiness as the dependent variable, rather than satisfaction.For the reasons described above, happiness seems a more appropriatedependent variable in this case than satisfaction.

METHOD

Sample

The data were obtained from five administrations ofGeneral SocialSurvey(GSS), a face-to-face interview survey conducted annually bythe National Opinion Research Center (Davis and Smith, 1991). Dataused here were from 1973, 1975, 1989, 1990 and 1991. The sixthdata set used here was reported first by Andrews and Withey (1974,1976) and collected in 1972. Survey participants were selected by amulti-stage, stratified, full probability sampling of English speakingpersons living in non-institutional quarters in the continental UnitedStates. A split-ballot design allowed all survey items to be asked of atleast one-third of all participants in the GSS surveys. Demographicdata for the samples are compared in Table I.

Our rationale for selecting these six data sets was as follows.The Andrews and Withey data set provided the strongest measuresof affect of all the data sets available (as described below). Yet wewere concerned that the age of this data set might make resultsderived from it invalid. We replicated the analyses of the Andrews

Page 4: Happiness and Satisfaction: Measures of Affect and Cognition?

198 KAREN J. CROOKER AND JANET P. NEAR

and Withey data set on five of the GSS surveys to assess reliability– first on the 1973 data set, which was most comparable to Andrewsand Withey in date of collection, and then to more recent GSS datasets to see whether results changed over time. It should be notedthat the use of existing data for secondary analyses represents adilemma: while the data collection techniques were superior to themethods that would be financially feasible for the typical researcher,there are instances where the variable measurement is not what wewould have hoped. Since our interest in these data differs from thatof its originators, this is certainly to be expected. Nonetheless, it isa frustration, as noted below in our discussion of the measures.

Measures

Dependent variable.The dependent variable, happiness, was mea-sured in all six data sets by the following single item with a threepoint scale: “Taken all together, how would you say things are thesedays – would you say that you are very happy [one], pretty happy[two], or not too happy [three]?” (see Appendix A for measures).This is one of two measures most commonly used in this coun-try; as Andrews and Robinson note, no other measure “has seemedmarkedly better” than these measures (1991: p. 71). While it wouldcertainly have been preferable to have more than one item to mea-sure this key concept, two important points should be noted. First,Scarpello and Campbell (1983) found in their study of job satis-faction measures that single item measures of satisfaction were asvalid and reliable as measures based on multiple items. Second,studies using the Andrews and Withey (1976) measure of life satis-faction, the average of a single item measure used twice during theinterview, have found a substantial (in the range of 0.70–0.80) test-retest reliability coefficient for this item (e.g., Headey, Holmstromand Wearing, 1984), suggesting again that the single item measureof satisfaction (and presumably, by extension, happiness) may beas useful as any multiple item measure. In any case, since we areengaging in secondary analysis of existing data, we do not have theluxury at this point of reconstructing the design of measures.

Average scores on the dependent variable were remarkably stable,over time, supporting Veenhoven’s (1984) contention that happinessvaries little through history (see Table II).

Page 5: Happiness and Satisfaction: Measures of Affect and Cognition?

HA

PP

INE

SS

AN

DS

AT

ISFA

CT

ION

199TABLE IIDescriptive statistiscs

Andrew & Withey General Social Survey

1972 1973 1975 1989 1990 1991

Mean S.D. N Mean S.D. N Mean S.D. N Mean S.D. N Mean S.D. N Mean S.D. N

Demographics:income 10.17 4.40 1046 7.65 3.15 1431 7.99 3.18 1448 13.63 4.98 1380 13.79 4.99 1229 13.62 5.32 1373gender 0.57 0.50 1072 0.53 0.50 1504 0.55 0.50 1490 0.57 0.50 1537 0.56 0.50 1372 0.58 0.49 1517marital status 0.33 0.47 1068 0.29 0.45 1504 0.33 0.47 1490 0.45 0.50 1537 0.47 0.50 1371 0.47 0.50 1517age 44.10 17.79 1068 44.18 16.75 1500 44.31 17.67 1485 45.44 17.81 1533 45.96 18.07 1372 45.63 17.81 1514education 50.39 21.60 1072 11.60 3.33 1499 11.69 3.12 1487 12.73 3.04 1530 12.82 3.10 1370 12.88 2.98 1510

Satisfaction with:job 2.69 1.20 698 1.67 0.80 1141 1.63 0.81 1165 1.73 0.82 1206 1.72 0.80 1041 1.74 0.80 1149community/city 2.78 1.17 1039 2.91 1.58 1502 2.88 1.58 1483 2.97 1.52 1028 2.90 1.43 898 2.91 1.45 1015friends 2.39 0.83 1045 2.23 1.28 1495 2.24 1.22 1484 2.16 1.21 1029 2.20 1.28 895 2.29 1.32 1016spare time/hobbies 2.89 1.12 1032 2.78 1.69 1487 2.73 1.63 1477 2.63 1.54 1026 2.65 1.53 892 2.70 1.53 1012family 2.30 1.05 918 2.11 1.39 1493 2.03 1.33 1482 2.10 1.38 1028 2.12 1.38 897 2.13 1.39 1012health 2.86 1.36 1051 2.60 1.57 1501 2.57 1.48 1483 2.61 1.46 1029 2.57 1.44 896 2.69 1.50 1014income/finances 3.45 1.43 1046 1.93 0.74 1501 1.96 0.76 1479 1.95 0.75 1532 1.97 0.75 1367 1.99 0.74 1506

Affect Scale 0.50 0.40 1503 0.48 0.39 1489 0.50 0.39 1035 0.51 0.38 899 0.52 0.39 1016Positive Affect 3.15 1.52 1030Negative Affect 1.61 1.50 1030Affect Balance 6.54 2.12 1030Total Affect 4.58 2.30 1072Happiness 2.79 1.12 1056 1.77 0.66 1500 1.80 0.65 1485 1.77 0.61 1526 1.76 0.60 1361 1.80 0.62 1504

Page 6: Happiness and Satisfaction: Measures of Affect and Cognition?

200 KAREN J. CROOKER AND JANET P. NEAR

Domain satisfactions.The independent satisfaction variables, rep-resenting domain satisfactions, were also measured by single items.For the GSS surveys, job satisfaction utilized a four point scale: “Onthe whole, how satisfied are you with the work you do – would yousay you are very satisfied [one], moderately satisfied [two], a lit-tle dissatisfied [three], or very dissatisfied [four]?” Satisfaction withfinancial situation utilized a three-point scale: “We are interested inhow people are getting along financially these days. So far as youand your family are concerned, would you say that you are prettywell satisfied [one] with your present financial situation, more orless satisfied [two], or not satisfied at all [three]?” The remainingsatisfaction items used seven point scales (one = a very great deal,two = a great deal, three = quite a bit, four = a fair amount, five= some, six = a little, seven = none) in response to the followingquestion: “For each area of life I am going to name, tell me thenumber that shows how muchsatisfactionyou get from that area”.The areas named included (a) “the city or place you live in”, (b)“your non-working activities – hobbies and so on”, (c) “your familylife”, (d) “your friendships”, (e) “your health and physical condi-tion”. In the Andrews and Withey data, all items were rated onthe seven-point “Delighted-Terrible” scale. Items took the form of“How do you feel about: : : your job/income/community/way youspend your spare time/your own family life/your friends/your ownhealth and physical condition”. The full text of these items is pro-vided in Appendix A. Following the general trend in this literature,noted above, we assumed that these measures tapped the cognitiveorientation of respondents, rather than their affective response tothese domains.

Affect.The remaining independent variable, affect, was developedby using principal components analysis with orthogonal rotation anditem trimming in data reduction. In the Andrews and Withey data setwe found ten items and we identified six items in the GSS data whichwe thought represented affect: that is, measures of how respondentsfelt about life, rather than their evaluation of it. This definition ofaffect has its roots in the work of Bradburn (1969) and follows a longtradition in the comparison of positive and negative affect measures(e.g., Diener, 1994; Headey et al., 1984) and discussion of differ-ences between measures of cognition versus affect (e.g., Organ and

Page 7: Happiness and Satisfaction: Measures of Affect and Cognition?

HAPPINESS AND SATISFACTION 201

Near, 1985). As noted above, we feel that the Andrews and Witheydata provided the affect measure with the greatest face validity.These items referred directly to how respondents characterized theirfeelings, which fell on two factors (Table III).

We believed that we could not rely on these data exclusively,however, because they are over 20 years old. We used more recentGSS data for this reason. We also used older GSS data sets to deter-mine whether the same relationships found in the older Andrews andWithey data would also hold in GSS data from the same time periodeven though different affect measures were used.

In any case, the six items identified were reduced to three bythe analysis, which resulted in a single factor (alpha = 0.67) asseen in Table III. These three items represent respondents’ positiveviews of other people as fair, helpful and to be trusted; two ofthem were obtained from Scheussler (1982) (the full text of theseitems is provided in Appendix B). The three items were standardizedand averaged to form a scale of positive affect. No correspondingmeasure of negative affect was available in these data. It should bereiterated that several studies have found that positive and negativeaffect are orthogonal (e.g., Diener and Emmons, 1985; Headey etal., 1984) at least under certain conditions (Diener and Iran-Nejad,1986).

Control variables.The following control variables were enteredfirst in all regression analyses: gender, age, years of education,family income and marital status. We entered these variables firstin the analyses because our goal was to assess incremental varianceexplained by affect after all other variables, especially cognitivemeasures of satisfaction, had been entered. Therefore, it seemedreasonable to also enter the demographics first, so that their effectswould be controlled statistically before we estimated the effects ofaffect. Of course, this is a conservative strategy: we wanted to knowwhat difference affect would make by itself, when all potentiallyconfounding variables had first been assessed.

Page 8: Happiness and Satisfaction: Measures of Affect and Cognition?

202K

AR

EN

J.CR

OO

KE

RA

ND

JAN

ET

P.N

EA

RTABLE IIIFactor analysis for affect items

Andrews & Withey General Social Surveys

1972 November 1973 1975 1989 1990 1991Item Factor 1 Factor 2 Factor Factor Factor Factor Factor

Positive affect:excited or interested 0.635proud of compliment 0.666pleased with accomplishment 0.759on top of the world 0.579thing going your way 0.674

Negative-affect:restless 0.620lonely or remote 0.715bored 0.691depressed or very unhappy 0.770upset because of criticism 0.485

Affect scale:fair 0.676 0.808 0.800 0.801 0.817helpful 0.632 0.776 0.769 0.783 0.767trusted 0.593 0.748 0.741 0.726 0.726

Page 9: Happiness and Satisfaction: Measures of Affect and Cognition?

HA

PP

INE

SS

AN

DS

AT

ISFA

CT

ION

203

TABLE IIIContinued

Andrews & Withey General Social Surveys

1972 November 1973 1975 1989 1990 1991Item Factor 1 Factor 2 Factor Factor Factor Factor Factor

Eigenvalue 2.317 2.247 1.901 1.815 1.899 1.926 1.899Percent of variance explained 23.2 22.5 63.4 60.5 31.7 21.1 31.7Reliability alpha 0.687 0.677 0.713 0.675 0.666 0.665 0.669

N 1031 1031 1485 1458 1020 885 994

Page 10: Happiness and Satisfaction: Measures of Affect and Cognition?

204 KAREN J. CROOKER AND JANET P. NEAR

RESULTS

Descriptive statistics are shown in Table II and correlations inTable IV, and they indicate that both the dependent and indepen-dent variables have adequate discriminant validity and should poseno multicollinearity difficulties for subsequent regression analyses.

The central issue of whether affect measures will explain signif-icant incremental variance in happiness, beyond that explained bycognitive variables, was assessed with hierarchical multiple regres-sion. All other independent variables – satisfaction with job, finances,location, hobbies, family, friends, and health were entered together inthe regression first. Then affect was entered alone in order to assessthe incremental variance explained (see Table V). Listwise deletionof cases was used in cases with missing data, resulting in use ofsmaller samples. While the cognitive variables explained between21 and 31 percent of the variance in happiness, affect did not addmuch additional variance (0.01 for the GSS data and .05 for theAndrews and Withey data, using adjusted variance explained).

DISCUSSION

Our expectation was that independent variables, including bothcognitive and affective measures, would explain greater variancein happiness, an allegedly affective measure, than would the usualcast of cognitive variables alone. In fact, this was not the case. Inprevious analyses using satisfaction and happiness measures as thedependent variables, similar results were obtained (McKennell et al.,1980; Schwarz and Clore, 1983). If, indeed, satisfaction representsa cognitive measure and happiness an affective measure, then wewould expect that independent variables that were affective in ori-entation would contribute more variance to the affective dependentvariable than would the cognitive independent variables.

It would seem that there are two possible explanations for ourresults. It is possible that the classification of variables as eithercognitive or affective in orientation is not as clear as we had initiallyexpected. That is, the discriminant validity of so-called cognitivemeasures, such as satisfaction, and alleged affective measures, suchas happiness, is in question. Instead, both measures may have a

Page 11: Happiness and Satisfaction: Measures of Affect and Cognition?

HA

PP

INE

SS

AN

DS

AT

ISFA

CT

ION

205TABLE IVCorrelations

Part 1

A & W 1972, part A 1 2 3 4 5 6 7 8 9

Demographics1 income 1.002 gender –0.07� 1.003 martial status –0.38�� 0.14�� 1.004 age –0.23�� 0.07� 0.12�� 1.005 education 0.39�� –0.10�� –0.08�� –0.38�� 1.00

Satisfaction6 job –0.05 –0.02 0.03 –0.15�� –0.02 1.007 community –0.09�� 0.01 0.08� –0.18�� 0.03 0.21�� 1.008 friends –0.05 –0.02 –0.02 –0.10�� –0.01 0.28�� 0.32�� 1.009 spare time –0.06 0.04 0.04 –0.04 –0.01 0.30�� 0.23�� 0.34�� 1.00

10 family –0.12�� 0.04 0.24�� 0.04 –0.07 0.21�� 0.25�� 0.32�� 0.36��

11 health –0.16�� 0.08� 0.08� 0.25�� –0.23�� 0.21�� 0.13�� 0.23�� 0.30��

12 income –0.28�� 0.06 0.17�� –0.09�� –0.10�� 0.38�� 0.25�� 0.24�� 0.35��

13 Positive affect 0.22�� 0.00 –0.12�� –0.17�� 0.26�� –0.20�� –0.08�� –0.018�� –0.25��

14 Negative affect –0.11�� 0.06 0.17�� –0.15�� –0.07� 0.25�� 0.14�� 0.11�� 0.23��

15 Total affect 0.10�� 0.04 0.01 –0.25�� 0.16�� 0.04 0.03 –0.06 0.0016 Affect balance 0.23�� –0.04 –0.21�� 0.02 0.23�� –0.32�� –0.16�� –0.21�� –0.34��

17 Happiness –0.19�� 0.04 0.17�� 0.03 –0.15�� 0.29�� 0.22�� 0.26�� 0.37��

Page 12: Happiness and Satisfaction: Measures of Affect and Cognition?

206K

AR

EN

J.CR

OO

KE

RA

ND

JAN

ET

P.N

EA

RTABLE IVContinued

Part 2A & W 1972, part B 10 11 12 13 14 15 16 17

Demographics

1 income2 gender3 martial status4 age5 education

Satisfaction6 job7 community8 friends9 spare time

10 family 1.0011 health 0.24�� 1.0012 income 0.35�� 0.25�� 1.0013 Positive affect –0.30�� –0.27�� –0.23�� 1.0014 Negative affect 0.21�� 0.22�� 0.26�� 0.01 1.0015 Total affect –0.08� –0.05 0.03 0.71�� 0.71�� 1.0016 Affect balance –0.36�� –0.35�� –0.34�� 0.71�� –0.70�� 0.01 1.0017 Happiness 0.42�� 0.29�� 0.40�� –0.39�� 0.31�� –0.04 –0.50�� 1.00

Page 13: Happiness and Satisfaction: Measures of Affect and Cognition?

HA

PP

INE

SS

AN

DS

AT

ISFA

CT

ION

207

TABLE IVContinued

Part 3GSS 1973, part A 1 2 3 4 5 6 7

Demographics1 income 1.002 gender –0.07�� 1.003 martial status –0.30�� 0.03 1.004 age –0.17�� –0.01 –0.02 1.005 education 0.42�� –0.03 –0.04 –0.31�� 1.00

Satisfaction6 job –0.14�� –0.05 –0.14�� –0.16�� –0.06 1.007 community 0.06� –0.05� 0.07�� –0.24�� 0.03 0.24�� 1.008 friends –0.13�� –0.06� 0.09�� –0.07� –0.11�� 0.22�� 0.31��

9 spare time –0.20�� 0.04 0.06� 0.01 –0.18�� 0.17�� 0.28��

10 family –0.22�� –0.01 0.40�� 0.05� –0.12�� 0.19�� 0.24��

11 health –0.23�� 0.05 0.09�� 0.20�� –0.22�� 0.18�� 0.17��

12 income –0.22�� –0.02 0.07�� –0.14�� –0.05� 0.19�� 0.28��

13 Affect –0.22�� –0.03 0.10�� –0.09�� –0.24�� 0.22�� 0.26��

14 Happiness –0.20�� –0.02 0.26�� –0.05 –0.05 0.25�� 0.29��

Page 14: Happiness and Satisfaction: Measures of Affect and Cognition?

208K

AR

EN

J.CR

OO

KE

RA

ND

JAN

ET

P.N

EA

R

TABLE IVContinued

Part 4GSS 1973, part B 8 9 10 11 12 13 14

Demographics1 income2 gender3 martial status4 age5 education

Satisfaction6 job7 community8 friends 1.009 spare time 0.33�� 1.00

10 family 0.43�� 0.31�� 1.0011 health 0.28�� 0.29�� 0.27�� 1.0012 income 0.16�� 0.15�� 0.18�� 0.15�� 1.0013 Affect 0.21�� 0.18�� 0.17�� 0.17�� 0.23�� 1.0014 Happiness 0.30�� 0.23�� 0.40�� 0.30�� 0.31�� 0.23�� 1.00

Page 15: Happiness and Satisfaction: Measures of Affect and Cognition?

HA

PP

INE

SS

AN

DS

AT

ISFA

CT

ION

209

TABLE IVContinued

Part 5GSS 1975, part A 1 2 3 4 5 6 7

Demographics1 income 1.002 gender –0.11�� 1.003 martial status –0.37�� 0.09�� 1.004 age –0.21�� 0.00 –0.02 1.005 education 0.38�� –0.03 –0.02 –0.35�� 1.00

Satisfaction6 job –0.13�� 0.04 0.05 –0.18�� –0.03 1.007 community –0.04 –0.01 0.10�� –0.21�� 0.03 0.24�� 1.008 friends –0.07� –0.06� 0.09�� 0.03 –0.10�� 0.14�� 0.28��

9 spare time –0.10�� 0.02 0.05 0.05 –0.17�� 0.13�� 0.24��

10 family –0.15�� –0.05 0.36�� 0.06� –0.02 0.15�� 0.24��

11 health –0.24�� 0.06� 0.10�� 0.23�� –0.22�� 0.08�� 0.17��

12 income –0.20�� 0.00 0.06� –0.19�� –0.07�� 0.23�� 0.18��

13 Affect –0.016�� –0.02 0.07�� –0.02 –0.27�� 0.10�� 0.15��

14 Happiness –0.14�� –0.01 0.21�� –0.07� –0.08�� 0.28�� 0.27��

Page 16: Happiness and Satisfaction: Measures of Affect and Cognition?

210K

AR

EN

J.CR

OO

KE

RA

ND

JAN

ET

P.N

EA

R

TABLE IVContinued

Part 6GSS 1975, part B 8 9 10 11 12 13 14

Demographics1 income2 gender3 marital status4 age5 education

Satisfaction6 job7 community8 friends 1.009 spare time 0.37�� 1.00

10 family 0.37�� 0.27�� 1.0011 health 0.28�� 0.27�� 0.25�� 1.0012 income 0.09�� 0.09�� 0.08�� 0.09�� 1.0013 Affect 0.18�� 0.14�� 0.10�� 0.15�� 0.15�� 1.0014 Happiness 0.24�� 0.23�� 0.31�� 0.24�� 0.27�� 0.16�� 1.00

Page 17: Happiness and Satisfaction: Measures of Affect and Cognition?

HA

PP

INE

SS

AN

DS

AT

ISFA

CT

ION

211

TABLE IVContinued

Part 7GSS 1989, part A 1 2 3 4 5 6 7

Demographics1 income 1.002 gender –0.16�� 1.003 marital status –0.34�� 0.12�� 1.004 age –0.16�� 0.07�� –0.09�� 1.005 education 0.44�� –0.10�� –0.02 –0.26�� 1.00

Satisfaction6 job –0.21�� –0.01 0.10�� –0.11�� –0.08�� 1.007 community –0.14�� –0.02 0.07� –0.25�� –0.03 0.17�� 1.008 friends –0.14�� –0.05 0.06 –0.06 –0.09�� 0.221�� 0.28��

9 spare time –0.24�� 0.09�� 0.04 0.03 –0.22�� 0.19�� 0.23��

10 family –0.20�� –0.04 0.33�� –0.02 0.00 0.18�� 0.23��

11 health –0.18�� 0.02 0.02 0.14�� –0.16�� 0.19�� 0.24��

12 income –0.33�� 0.00 0.11�� –0.17�� –0.15�� 0.30�� 0.22��

13 Affect –0.25�� –0.06 0.11�� –0.17�� –0.23�� 0.14�� 0.27��

14 Happiness –0.21�� 0.00 0.17�� –0.02 –0.06� 0.34�� 0.27��

Page 18: Happiness and Satisfaction: Measures of Affect and Cognition?

212K

AR

EN

J.CR

OO

KE

RA

ND

JAN

ET

P.N

EA

R

TABLE IVContinued

Part 8GSS 1989, part B 8 9 10 11 12 13 14

Demographics1 income2 gender3 marital status4 age5 education

Satisfaction6 job7 community8 friends 1.009 spare time 0.39�� 1.00

10 family 0.42�� 0.32�� 1.0011 health 0.36�� 0.30�� 0.32�� 1.0012 income 0.16�� 0.12�� 0.13�� 0.17�� 1.0013 Affect 0.18�� 0.18�� 0.13�� 0.10�� 0.16�� 1.0014 Happiness 0.35�� 0.26�� 0.31�� 0.31�� 0.26�� 0.13�� 1.00

Page 19: Happiness and Satisfaction: Measures of Affect and Cognition?

HA

PP

INE

SS

AN

DS

AT

ISFA

CT

ION

213

TABLE IVContinued

Part 9GSS 1990, part A 1 2 3 4 5 6 7

Demographics1 income 1.002 gender –0.11�� 1.003 marital status –0.42�� 0.12�� 1.004 age –0.14�� 0.07� 0.00 1.005 education 0.37�� –0.02 –0.03 –0.17�� 1.00

Satisfaction6 job –0.16�� –0.01 0.09�� –0.12�� –0.08� 1.007 community –0.13�� –0.04 0.10�� –0.16�� –0.08� 0.25�� 1.008 friends –0.08� –0.08� 0.08� 0.02 –0.13�� 0.21�� 0.35��

9 spare time –0.20�� 0.03 0.09�� 0.05 –0.21�� 0.18�� 0.36��

10 family –0.21�� –0.03 0.37�� 0.09� 0.09�� 0.20�� 0.31��

11 health –0.22�� 0.04 0.12�� 0.18�� –0.20�� 0.22�� 0.29��

12 income –0.26�� 0.04 0.09�� –0.16�� –0.10�� 0.33�� 0.25��

13 Affect –0.15�� –0.01 0.10�� –0.17�� –0.23�� 0.13�� 0.21��

14 Happiness –0.17�� 0.01 0.28�� 0.03 –0.08�� 0.32�� 0.26��

Page 20: Happiness and Satisfaction: Measures of Affect and Cognition?

214K

AR

EN

J.CR

OO

KE

RA

ND

JAN

ET

P.N

EA

R

TABLE IVContinued

Part 10GSS 1990, part B 8 9 10 11 12 13 14

Demographics1 income2 gender3 marital status4 age5 education

Satisfaction6 job7 community8 friends 1.009 spare time 0.40�� 1.00

10 family 0.49�� 0.33�� 1.0011 family 0.41�� 0.37�� 0.39�� 1.0012 health 0.10�� 0.15�� 0.13�� 0.15�� 1.0013 Affect 0.23�� 0.22�� 0.19�� 0.14�� 0.21�� 1.0014 Happiness 0.12�� 0.29�� 0.43�� 0.33�� 0.29�� 0.17�� 1.00

Page 21: Happiness and Satisfaction: Measures of Affect and Cognition?

HA

PP

INE

SS

AN

DS

AT

ISFA

CT

ION

215

TABLE IVContinued

Part 11GSS 1991, part A 1 2 3 4 5 6 7

Demographics1 income 1.002 gender –0.13�� 1.003 marital status –0.38�� 0.09�� 1.004 age –0.02 0.07�� –0.04 1.005 education 0.32�� –0.10�� –0.05� –0.25�� 1.00

Satisfaction6 job –0.17�� 0.09�� 0.08�� –0.09�� –0.08�� 1.007 community –0.10�� –0.03 0.06 –0.20�� –0.07� 0.17�� 1.008 friends –0.14�� –0.03 0.07 –0.02 –0.14�� 0.19�� 0.30��

9 spare time –0.14�� 0.06 0.01 0.01 –0.18�� 0.20�� 0.33��

10 family –0.20�� –0.05 0.32�� 0.04 –0.09�� 0.19�� 0.26��

11 health –0.23�� 0.08�� 0.06� 0.16�� –0.18�� 0.19�� 0.20��

12 income –0.23�� 0.07�� 0.12�� –0.23�� –0.09�� 0.22�� 0.22��

13 Affect –0.15�� –0.06 0.08� –0.14�� –0.23�� 0.16�� 0.21��

14 Happiness –0.21�� 0.06� 0.18�� –0.04 –0.12�� 0.31�� 0.28��

Page 22: Happiness and Satisfaction: Measures of Affect and Cognition?

216K

AR

EN

J.CR

OO

KE

RA

ND

JAN

ET

P.N

EA

R

TABLE IVContinued

Part 12GSS 1991, part B 8 9 10 11 12 13 14

Demographics1 income2 gender3 marital status4 age5 education

Satisfaction6 job7 community8 friends 1.009 spare time 0.44�� 1.00

10 family 0.48�� 0.34�� 1.0011 health 0.37�� 0.30�� 0.31�� 1.0012 income 0.19�� 0.20�� 0.16�� 0.16�� 1.0013 Affect 0.21�� 0.17�� 0.16�� 0.06� 0.18�� 1.0014 Happiness 0.30�� 0.29�� 0.15�� 0.26�� 0.30�� 0.19�� 1.00

� = significance� 0.05 �� = significance� 0.01

Page 23: Happiness and Satisfaction: Measures of Affect and Cognition?

HA

PP

INE

SS

AN

DS

AT

ISFA

CT

ION

217

TABLE VSummary of hierarchical regression for variables predicting happiness

Independent Variables Andrews & Withey General Social Surveys1972 November 1973 1975 1989 1990 1991

b t p b t p b t p b t p b t p b t p

Demographics:income –0.023 –1.855+ –0.087 –0.037 –4.938��� –0.166 –0.01 –2.476� –0.088 –0.026 –4.430��� –0.200 –0.016 –2.67�� –0.121 –0.020 –3.742��� –0.165gender –0.700 –0.776 –0.033 –0.076 –1.978� –0.059 –0.06 –1.559 –0.047 –0.057 –1.253 –0.047 –0.007 –0.16 –0.006 0.028 0.589 0.023marital status 0.209 1.696+ 0.077 0.290 6.265�� 0.193 0.21 4.719��� 0.154 0.114 2.304� 0.093 0.353 6.83��� 0.292 0.072 1.412 0.059age –0.003 –0.932 –0.041 –0.002 –1.085 –0.033 –0.00 –2.291� –0.071 –0.002 –1.032 –0.039 0.002 1.19 0.046 0.000 0.258 0.010education –0.002 –1.040 –0.048 –0.00 –0.141 –0.005 –0.02 –3.036�� –0.101 0.011 1.307 0.055 –0.004 –0.40 –0.017 –0.011 –1.192 –0.049

Change in R Squared 0.023� 0.086��� 0.058��� 0.057��� 0.12��� 0.048���

Adjusted R Squared 0.015 0.082 0.054 0.051 0.12 0.041

Satisgaction with:job 0.065 1.844+ 0.075 0.073 3.184��� 0.089 0.15 6.751��� 0.194 0.154 5.894��� 0.207 0.129 4.77��� 0.169 0.126 4.707��� 0.184community/city 0.024 0.645 0.026 0.032 2.611�� 0.778 0.03 2.583�� 0.077 0.042 2.932�� 0.104 0.022 1.32 0.042 0.060 3.832��� 0.130friends 0.071 1.238 0.511 0.061 3.720��� 0.117 0.02 2.672+ 0.052 0.077 3.752��� 0.149 0.043 2.13� 0.071 0.031 1.610 0.038spare time/hobbies 0.101 2.334� 0.101 0.011 0.946 0.029 2.952�� 0.089 0.034 2.121� 0.081 0.029 1.87+ 0.076 0.021 1.260 0.064family 0.301 7.104��� 0.300 0.073 4.490��� 0.148 0.04 2.621�� 0.082 0.050 2.634�� 0.109 0.084 4.23��� 0.263 0.075 3.956��� 0.176health 0.084 2.394� 0.095 0.062 4.773��� 0.139 0.05 3.906��� 0.116 0.028 1.723+ 0.064 0.044 2.49�� 0.100 0.036 2.163� 0.095income/finances 0.113 3.411��� 0.147 0.151 6.050��� 0.172 0.13 5.602��� 0.160 0.092 3.130�� ‘0.114 0.121 4.21��� 0.174 0.115 3.775��� 0.150

Change in R Squared 0.267��� 0.187��� 0.165��� 0.200��� 0.22��� 0.215���

Adjusted R Sqaured 0.275 0.265 0.215 0.245 0.34 0.250

Page 24: Happiness and Satisfaction: Measures of Affect and Cognition?

218K

AR

EN

J.CR

OO

KE

RA

ND

JAN

ET

P.N

EA

R

TABLE VContinued

Independent Variables Andrews & Withey General Social Surveys1972 November 1973 1975 1989 1990 1991

b t p b t p b t p b t p b t p b t p

Affect Scale 0.162 3.435��� 0.100 0.04 0.855 0.024 0.067 1.169 0.042 0.011 0.18 0.007 0.089 1.525 0.056Change in R Sqaured 0.008��� 0.001 0.001 0.00 0.003Adjusted R Squared 0.272 0.215 0.245 0.34 0.251

Positive Affect –0.126 –4.595��� –0.178Negative Affect 0.112 4.060��� 0.156Change in R Squared 0.040���

Adjusted R Squared 0.313

Affect Balance –0.119 –5.705��� –0.233Change in R Squared 0.040���

Adjusted R Squared 0.314

Total Affect 0.000 0.030 0.001Change in R Squared 0.000Adjusted R Squared 0.272

N 575 1114 1121 787 66 745

���2-tailed t test is significant, p< 0.001 ��2-tailed t test is significant, p< 0.01 +2-tailed t test is significant, p<

0.10 ���F test is significant, p< 0.001

Page 25: Happiness and Satisfaction: Measures of Affect and Cognition?

HAPPINESS AND SATISFACTION 219

component that is cognitive and a component that is affective. Thispossibility is suggested by Brief and Roberson’s (1989) finding thatsome job satisfaction scales appear to have affective componentswhile others are primarily cognitive. Or, as Diener (1994) suggests,perhaps both are components of a higher-order construct.

Alternatively, limitations in the measurement of either our depen-dent variable, happiness, or the independent variable, affect, mayhave reduced their utility in this analysis. Certainly, it would havebeen preferable to have a multiple item measure of happiness,although it is not clear that such a measure would have been psycho-metrically superior to the current measure (Andrews and Robinson,1991). Likewise, we would have preferred to use a more standardmeasure of affect; nonetheless, the scale used here met the originalcriterion in that it clearly referred to feelings rather than evalua-tions and it was reliable. It is interesting that the adjusted varianceexplained by the affect scale in the Andrews and Withey data wasslightly higher (0.05) than that explained by the affect scale usedwith the GSS samples (0.00), perhaps indicating greater validity ofthe Andrews and Withey scale than of the GSS scale. On the otherhand, satisfaction measures preceded the happiness measures in alldata sets except that from Andrews and Withey, where affect mea-sures preceded the happiness measure. As Schwarz (1994) argues,happiness measures may show artificially high correlations withmeasures that precede them, and it may be that the affect measuresin the Andrews and Withey data were more strongly related to hap-piness than in the other data sets simply because they were linkedin respondents’ minds. Nonetheless, neither scale contributed muchadditional variance to that explained by the cognitive independentvariables (0.20–0.30). If cognitive and affective measures were trulyindependent, we would have expected a greater portion of uniquevariance in an affective dependent variable (i.e., happiness) to beexplained by affective independent variables than by cognitive inde-pendent variables.

Conclusions

Our results suggest the need for empirical reevaluation of the fre-quently cited view that satisfaction is a cognition and happiness anaffect. Studies of job satisfaction and life satisfaction alike have

Page 26: Happiness and Satisfaction: Measures of Affect and Cognition?

220 KAREN J. CROOKER AND JANET P. NEAR

been criticized for their neglect of affect measures and exclusivereliance on cognitive measures (Organ and Near, 1985). Yet, thehypothesis that affective measures produce different results thancognitive results has not been rigorously tested; our investigationprovides preliminary evidence that the two constructs may not differ.Continued reliance on the truism that they do differ, without morerigorous empirical examination of its validity, is counterproductiveto our increased understanding of the relationship between subjec-tive well being and other domains of life, such as job, family andleisure.

REFERENCES

Andrews, F.M. and A.C. McKennell: 1980. ‘Measures of self-reported well-being:Their affective, cognitive, and other components’, Social Indicators Research 8,pp. 127–155.

Andrews, F.M. and J.P. Robinson: 1991. ‘Measures of subjective well-being’, inJ.P. Robinson, P.R. Shaver and L.S. Wrightsman (eds.), Measures of Personalityand Social Psychological Attitudes, Vol. 1 ( Academic Press, Inc., San Diego),pp. 61–76.

Andrews, F.M. and S.B. Withey: 1976. Social Indicators of Well-being (PlenumPress, New York).

Andrews, F.M. and S.B. Withey: 1974. ‘Developing measures of perceived lifequality: Results from several national surveys’, Social Indicators Research 1,pp. 1–26.

Bradburn, N.M.: 1969. The Structure of Psychological Well Being (Aldine,Chicago).

Brief, A.P. and L. Roberson: 1989. ‘Job attitude organization: An exploratorystudy’, Journal of Applied Social Psychology 19, pp. 717–727.

Brief, A.P., A.H. Butcher and L. Roberson: 1995. ‘Cookies, disposition, and jobattitudes: The effects of positive mood-inducing events and negative affectivityon job satisfaction in a field experiment’, Organizational Behavior and HumanDecision Processes 62, pp. 55–62.

Campbell, A.: 1976. ‘Subjective measures of well-being’, American Psychologist31, pp. 117–124.

Clark, L.A. and D. Watson: 1991. ‘General affective dispositions in physical andpsychological health’, in C.R. Snyder and D.R. Forsyth (eds.), Handbook ofsocial and clinical psychology (Pergamon Press, New York), pp. 221–245.

Davis, J.A. and T.W. Smith: 1991. General Social Surveys, 1972–1991 [machine-readable data file]. Principal Investigator, James A. Davis; Director and Co-Principal Investigator, Tom W. Smith. NORC ed. Chicago: National OpinionResearch Center, producer, 1991; Storrs, CT: The Roper Center for PublicOpinion Research, University of Connecticut, distributor. 1 data file (27,782logical records) and 1 codebook (989 p).

Diener, E.: 1984. ‘Subjective well-being’, Psychological Bulletin 95, pp. 542–575.

Page 27: Happiness and Satisfaction: Measures of Affect and Cognition?

HAPPINESS AND SATISFACTION 221

Diener, E.: 1994. ‘Assessing subjective well-being: Progress and opportunities’,Social Indicators Research 31, pp. 103–157.

Diener, E. and R.A. Emmons: 1985. ‘The independence of positive and negativeaffect’, Journal of Personality and Social Psychology 47, pp. 1105–1117.

Diener, E. and A. Iran-Nejad: 1986. ‘The relationship in experience betweenvarious types of affect’, Journal of Personality and Social Psychology 50,pp. 1031–1038.

Headey, B., E. Holmstrom and A. Wearing: 1984. ‘Well-being and ill-being:Different dimensions?’, Social Indicators Research 14, pp. 115–139.

Judge, T.A. and A. Eretz: 1995. Dispositional Influences on Enhancement andDetraction in the Reporting of Job and Life Affect. Unpublished working paper,Cornell University, New York State School of Industrial and Labor Relations,Ithaca.

McKennell, A.C.: 1978. ‘Cognition and affect in perceptions of well-being’, SocialIndicators Research 5, pp. 389–426.

McKennell, A.C. and F.M. Andrews: 1980. ‘Models of cognition and affect inperceptions of well-being’, Social Indicators Research 8, pp. 257–298.

McKennell, A., T. Atkinson and F.M. Andrews: 1980. ‘Structural constancies insurveys of perceived well-being’, in A. Szalai and F.M. Andrews (eds.), TheQuality of Life: Comparative Studies (Sage, London), pp. 111–128,

Organ, D.W. and J.P. Near: 1985. ‘Cognition vs. affect in measures of job satis-faction’, International Journal of Psychology 20, pp. 241–253.

Rain, J.S., I.M. Lane and D.D. Steiner: 1991. ‘A current look at the job satisfac-tion/life satisfaction relationship’, Human Relations 44, pp. 287–307.

Rice, R.W., J.P. Near and R.G. Hunt: 1980. ‘The job satisfaction-life satisfac-tion relationship: A review of empirical research’, Basic and Applied SocialPsychology 1, pp. 37–64.

Scarpello, V. and J.P. Campbell: 1983. ‘Job satisfaction: Are all the parts there?’,Personnel Psychology 36, pp. 577–600.

Scheussler, K.: 1982. Measuring Social Life Feelings (Jossey-Bass, San Francis-co).

Schwarz, N. and G.L. Clore: 1983. ‘Mood, misattribution, and judgements ofwell-being: informative and directive functions of affective states’, Journal ofPersonality and Social Psychology 45, pp. 512–523.

Schwarz, N.: 1994. ‘Judgment in a social context: Biases, shortcomings, and thelogic of conversation’, in M. Zanna (ed.), Advances in Experimental SocialPsychology, Vol. 20 (Academic Press, San Diego).

Veenhoven, R.: 1984. Conditions of Happiness (D. Reidel, Dordrecht, Holland).Wheeler, R.J.: 1991. ‘The theoretical and empirical structure of general well-

being’, Social Indicators Research 24, pp. 71–79.

University of Wisconsin Milwaukee Karen J. Crooker

Indiana University Janet P. Near

Page 28: Happiness and Satisfaction: Measures of Affect and Cognition?

222 KAREN J. CROOKER AND JANET P. NEAR

APPENDIX ADEPENDENT VARIABLES AND DOMAIN SATISFACTION ITEMS

TEXT OF QUESTIONS AS ASKED IN SURVEYS

Andrews and Withey, 1972, November

Taking all things together, how would you say things are these days– would you say you’re very happy, pretty happy, or not too happythese days?

1. very happy 3. pretty happy 5. not too happy

<how do you feel about> your job?<how do you feel about> the income you (and your family) have?<how do you feel about> this community as a place to live?<how do you feel about> the way you spend your spare time, yournon-working activities?<how do you feel about> your own family life – your wife/husband,your marriage, your children if any?<how do you feel about> your friends?<how do you feel about> your own health and physical condition?

1. delighted 5. mostly dissatisfied

2. pleased 6. unhappy3. mostly satisfied 7. terrible

4. mixed (about equally satisfied and dissatisfied)

General Social Survey, 1973, 1975, 1989, 1990, 1991

Taken all things together, how would you say things are these days– would you say that you are very happy, pretty happy, or not toohappy?

1. very happy 3. pretty happy 5. not too happy

On the whole, how satisfied are you with the work you do – wouldyou say you are very satisfied, moderately satisfied, a little dissatis-fied, or very dissatisfied?

1. very satisfied 3. a little dissatisfied

2. moderately satisfied 4. very dissatisfied

We are interested in how people are getting along financially thesedays. So far as you and your family are concerned, would you say

Page 29: Happiness and Satisfaction: Measures of Affect and Cognition?

HAPPINESS AND SATISFACTION 223

that you are pretty well satisfied with your present financial situation,more or less satisfied, or not satisfied at all?

1. pretty well satisfied 3. not satisfied at all

2. more or lesssatisfied

For each area of life I am going to name, tell me the number thatshows how muchsatisfactionyou get from that area.

the city or place you live inyour non-working activities – hobbies and so onyour family lifeyour friendshipsyour health and physical condition

1. a very great deal 5. some

2. a great deal 6. a little3. quiet a bit 7. none

4. a fair amount

APPENDIX BAFFECT ITEMS

TEXT OF QUESTIONS AS ASKED IN SURVEYS

Andress and Withey, 1972, November

Now I have some questions about how you have been feeling recent-ly. You can just answer “yes” or “no”. During the past few weeksdid you ever fell< >? How often during the past few weeks didyou feel< >? Was it just once, several times, or a lot of times?

particularly excited or interested in somethingso restless that you couldn’t sit long in a chairproud because someone complimented you on something youhad donevery lonely or remote from other peoplepleased about having accomplished somethingboredon top of the worlddepressed or very unhappy

Page 30: Happiness and Satisfaction: Measures of Affect and Cognition?

224 KAREN J. CROOKER AND JANET P. NEAR

that things were going your wayupset because someone criticized you

0. no never1. yes; once2. yes; several times3. yes; a lot of times4. yes; don’t know how many times

General Social Survey, 1973, 1975, 1989, 1990, 1991

Would you say that most of the time people try to be helpful, or thatthey are mostly just looking out for themselves?

1. try to be helpful2. just look out for themselves3. depends

Do you think most people would try to take advantage of you if theygot a chance, or would they try to be fair?

1. would take advantage2. would try to be fair3. depends

Generally speaking, would you say that most people can be trustedor that you can’t be too careful in dealing with people?

1. most people can be trusted2. can’t be too careful3. depends