(best) 55.the structure of psychological well-being revisited

8/6/2019 (Best) 55.the Structure of Psychological Well-being Revisited

1/9

Journal of Personality and Social Psychology1995, Vol. 69 , No. 4,719-727Copyright 1995 by the American ^ i ^ 0 1 ^ * ^ ! ^ ] ^

The St ructu re of Psychological Well-Being RevisitedCarol D . Ryff and Corey Lee M . Keyes

University of WisconsinMadisonA theoretical model of psychological well-being that encompasses 6 distinct dime nsions of wellness(Autonomy, Environm ental M astery, Personal Growth, Positive Relations With Others, Purpose inLife, Self-Acceptance) was tested with d ata from a nationally representative sam ple of adults (iV -1,108), aged 25 and older, who participated in telephone interviews. Confirmatory factor analysesprovided support for the proposed 6-factor model, with a single second-order super factor. T he m odelwas superior in fit over single-factor and other artifactual models. Age and sex differences on thevarious well-being dimensions replicated prior findings. Comparisons with other frequently usedindicators (positive and negative affect, life satisfaction) demonstrated that the latter neglect keyaspects of positive functioning emphasized in theories of health and well-being.

For more th an 20 years, the study of psychological well-beinghas been guided by two primary conceptions of positive func-tioning. One formulation, traceable to B radburn's (196 9) sem-inal work, distinguished between positive and negative affectand denned happiness as the balance between the two. Concep-tual and methodological refinements built on this early opera-tionalization of well-being. For example, the postulated inde-pendence of positive and negative affect was challenged andlinked with the failure to distinguish between the intensity andthe frequency of affect (Diener, Larsen, Levine, & Emmons,1985). Frequency of positive and negative affect tends to corre-late negatively, whereas intensity correlations are generally pos-itive. These conflicting relations were said to suppress the asso-ciation between positive and negative affect, thereby creatingan illusion that the components are independent. Of the two,frequency has been promoted as the better indicator of well-being because it can be better measured and is more stronglyrelated to long-term emotional well-being than intensity is(Diener & Larsen, 1993; Diener, Sandvik, & Pavot, 1991).Other initiatives have focused on measurement issues, callingfor more valid and reliable indicators of positive and negativeaffect (Watson, Clark, & Tellegen, 1988) and suggesting thatmeasureme nt error obscures the bipolarity of positive and neg-ative affect (G reen, G oldma n, & Salovey, 1 993).

The second primary conception, which has gained promi-nence among sociologists, emphasizes life satisfaction as the keyindicator of well-being. Viewed as a cognitive component, life

Carol D. Ryff, Department of Psychology, University of WisconsinMadison; Corey Lee M . Keyes, D epartm ent of Sociology, University ofWisconsinMadison.This research was supported by the John D. and Catherine T.MacArthur Foundation Research Network on Successful MidlifeDevelopment.Correspondence concerning this article should be addressed to CarolD. Ryff, Dep artme nt of Psychology, Brodgen Psychology Building, Uni-versity of Wisconsin, 1202 West Johnson Street, Madison, Wisconsin,53706, or to Carol D. Ryff, Institute on Aging, 2245 Medical SciencesCenter, University of Wisconsin Madison , 1300 University Avenue,Madison, Wisconsin 53706.

satisfaction was seen to complement happiness, the more affec-tive dimension of positive functioning (e.g., Andrews &McKennell, 1980; Andrews & Withey, 1976 ; Bryant & Veroff,1982; Cam pbell, Converse, & Rodgers, 197 6 ). Still other stud-ies parsed well-being according to global q uestions about overalllife satisfaction and domain-specific questions about work, in-come, social relationships, and neighborhood (Andrews, 1991;Diener, 1984). Interest in these investigations frequently cen-tered on social changewhether quality of life in Americameant something different from one era to the next and whetherreported levels of well-being and their correlates varied overtime (see also Bryant & Veroff, 1982).Altogether, prio r endeavors have grappled minimally with thecore underlying question: What does it mean to be well psycho-

logically? Tha t is, extant indicators have been perpetuated withlittle debate as to whether they captured key features of hum anwellness. Bradb urn's (19 6 9) classic study, for exam ple, gave lit-tle attention to the fundamental meaning of well-being. Thatpositive and negative affect emerged as independent dimensionswas, in fact, a serendipitousfindingfrom a study conceived forother purposes. Similarly, life satisfaction measures were gener-ated with a concern for practical ap plications of research find-ings, not explication of essential meanings of wellness (Sauer &Warland, 1982). Quality-of-life research has also been de-scribed as being data driven rather than based on a clear con-ceptual framework (Headey, Kelley, & Wearing, 1993). Water-man's (1993) distinction between eudaimonic and hedonicconceptions of happiness provides a notable exception in thislargely atheoretical climate.1The absence of theory-based formulations of well-being is

1 When it comes to the task of explaining the process of well-being,that is, how individuals come to possess or no t possess this quality, thetheoretical terrain is much richer (see Diener, 1984, for a review). Ourown work has targeted a dults' life experiences and their interpretationsof them (e.g., through social comparison processes) as key explanatoryfactors (e.g., Heidrich & Ryff, 1993a; Ryff & Essex, 1992; Ryff, Lee,Essex, & Schmutte, 1994). Here we emphasize the need for theory inthe prior task, namely, denning the essential nature of well-being, theindicators of which serve as outcome variables in such process studies.719


2/9

720 CAROL D. RYFF AND COREY LEE M. KEYESpuzzling given abundant accounts of positive functioning insubfields of psychology (see Ryff, 1985, 19 89a). From develop-mental psychology, Erikson's (1959) psychosocial stages, Buhl-er's (1935) basic life tendencies, and Neugarten's (1973) per-sonality changes articulate wellness as trajectories of continuedgrowth ac ross the life cycle. Clinical psychologists offer furtherdescriptions of well-being through Maslow's (19 6 8) conceptionof self-actualization, Allport's (196 1) formulation of maturity,Rogers' (1961) depiction of the fully functioning person, andJung's (1933) account of individuation. The mental health lit-erature, which typically elaborates the negative end of psycho-logical functioning, nonetheless includes some exposition ofpositive health (Birren& Renn er, 1980;Jahoda, 1958).The convergence of these multiple frameworks of positivefunctioning served as the theoretical foundation to generate amultidimensional model of well-being (Ryff, 1989b, 1995). In-cluded are six distinct components of positive psychologicalfunctioning (see App endix). In co mbination , these dimensionsencompass a brea dth of wellness that includes positive evalua-tions of oneself and one's past life (Self-Accep tance), a sense ofcontinued growth and development as a person (PersonalGro wth ), the belief that on e's life is purposeful and meaningful(Purpose in L ife), the possession of quality relations with others(Positive Relations With Others), the capacity to manageeffectively one's life and surrounding world (EnvironmentalMaste ry), and a sense of self-determination (Auton omy ).To understand the nature of wellness, descriptive studies havefocused on age and gender profiles. The original validation sam-ple (Ryff, 1989b) compared young (18-29 years old), midlife(30-64 years old), and old-aged (65 years old or older) adultsand found increm ental age profiles for E nvironmental Masteryand Autonomy (particularly from young adulthood to m idlife),decremental age profiles for Purpose in Life and PersonalGrowth (particularly from midlife to old age), and no agedifferences for Self-Acceptance and Positive Relations WithOthers. Most of these patterns were replicated in another study(Ryff, 1991) involving the same three age groups. In both in-vestigations, women scored significantly higher than men onPositive Relations W ith Others and Personal Grow th (findingsfor the latter dimension approached statistical significance inthe first study), with subsequent studies replicating these sexdifferences (Ryff et al., 1994; Ryff, Lee, & Na, 1993).

The proposed multidimensional structure of well-being hasnot, however, been investigated with analytic procedures thattest the fit of the theoretical model with empirical data. Highcorrelations among certain aspects of well-being (Ryff, 1989b)underscore the need for theory-guided structural analyses. Inaddition, the proposed theoretical m odel has not been assessedin a nationally representative sampleprior work has beenconducted primarily with selective, community samples. Datafrom representative samples are needed to test the generaliz-ability of prior patterns of age and sex differences. Thus, theobjectives of this study were threefold: (a ) to test, with a nation -ally representative sample, the proposed multidimensionalmodel of well-being; (b ) t o exa mine the replicative consistencyof age and sex differences on the various indicators of well-be-ing; and (c) to compare the relationships between the theory-based dimensions of well-being and three prominent indica-

tors from prior research (i.e., happiness, life satisfaction,depression).Method

SampleData are reported from a national probability sample of noninstitu-tionalized, English-speaking adults, aged 25 or older, residing in the 48contiguous states in the United States, and whose households includedat least one telephone. H ouseholds were selected with random digit di-aling proced ures. An adult from ea ch household was then selected ran-domly and interviewed for approximately 30 min by telephone. Theresponse rate was approximately 62%. Data were weighted to correctfor overrepresentation of households with more than one telephone line,for underrepresentation of adults aged 25 or older living in householdswith more than one adult, and to m atch census bureau estimates of theproportion of English-speaking adults, aged 25 or older, residing in themajor geographical regions (i.e., Northea st, Midwest, South, and West)of the United States.The total sample size was 1,108, of which 59% were female, 87% wereCaucasian, and the average age was 45.6 years (SD - 14.8 yea rs). Mostrespondents, about 70%, were married. Fully one-third had graduatedfrom high school only; 26% had some college background; and 16%were college graduates. Just over half of the sample repo rted an annualhousehold income between $10,000 and $39,999, of which 19% re-ported an income between $ 10,000 and $ 19,999,21 % between $20,000and $29,999, and 18% reported a household income of between $30,000and $39,999.For analytic purposes, we divided respondents into thre e age groups:young adults (n = 133) were between the ages of 25 and 29, midlifeadults (n = 805) were between the ages of 30 and 64, and older adults(n = 160) were 65 or older. About 60% of the sample in each age groupwas female. Though each age group consisted mostly of Caucasians, theracial homogeneity increased in the group of older adults. More thanhalf the young adults either had some college or graduated from college,whereas only 44% of the middle-aged, and still fewer of older adults(33%), had some college or graduated from college. Compared w ith theyoung adults, about 70% of the m iddle-aged, as well as the older adults,were married. The modal income range for young adults was between$30,000 and $39,999, $20,000-$29,999 for middle-aged adults, and$ 10,000-$ 19,999 for the older adults.Data from two additional samples (Ryff, 1989b; Ryffet al., 1994) areincluded for comparative purposes. Both were community volunteersamples, with the former consisting of young, midlife, and old-agedadults (see previous definitions of age ranges) and the latter includingonly midlife adults. D etailed descriptions of these samples a re availablein Ryff( 1989b) and Ryffet al. (1994).

Measures and ProcedureWe generated definitions of the six dimensions of psychological well-being (see Table 1) from the multiple theoretical accounts of positivefunctioning. In the initial validation study (Ryff, 1989b), each dimen-sion was operationalized with a 20-item scale (tha t showed high internalconsistency and test-rete st reliability as well as convergent and discrim -inant validity with other mea sures). To accommodate time and costrestrictions of a national survey, we chose only 3 of the original 20 itemsto me asure each construct. Because all parent scales had multifactorialstructures, we selected item s from subfactors w ithin each longer scale tomaximize the conceptual breadth of the shortened scales. The short-ened scales correlated from .70 to .89 with 20-item parent scales. Eachscale included both positively and negatively phrased items. The re-sponse scale was a 6-point continuum, ranging from completely dis-agree to completely agree. Interviewers administered the items using an


3/9

STRUCTURE OF WELL-BEING 721Table 1Descriptive Statistics of Theoretically Grounded Scales of Psychological Weil-Being

Scale1. Self-Acceptance2. Environmental Mastery3. Positive R elations4. Purpose in Life5. Personal Growth6. AutonomyaMSD

.5214.63.1

.46

.4914.92.8

.40.38

.5614.83.2

.22.13.14

.3314.43.2

.18.23.16.31.4015.72.5

.22.28.15.13.20.3715.22.6

Note. All correlation coefficients are statistically significant at the .05 level.

unfolding technique (Groves, 1989)first asking whether the respon-dent agreed or disagreed with the statem ent and then asking how much(strongly, moderately, or slightly).For comparative purposes, the interview also included single-item in-dicators of happiness and life satisfaction. For the former, respondentsanswered how much of the time during the past month (0 = none, 1 =some, 2 = m ost, 3 = all) they felt happy (me an score = 1.7, SD = 0.64).For the latter, respondents sum marized how things were going in theirlife on a scale from 0 (the worst possible life yo u could imagine) to 10(the best possible life you could imagine; mean score = 7.7, SZ> = 1.6).Depression was also measured w ith eight items indicating how much ofthe time (0 = none, 1 = some, 2 = m ost, 3 = all) during the past monthrespondents felt (1) full of life, (2) worn out, (3) tired, (4) downheartedand blue, (5) calm and peaceful, (6) nervous, (7) had a lot of energy,and (8) were so down in the dumps that nothing could cheer you up(mean score = 8.1, SD = 3.4). Factor analyses revealed twounderlyingdimensions, which were labeled Dysfunctional Energy and Dysfunc-tional Affect.From the two prior investigations (Ryff, 1989b; Ryffet al., 1994),data are reported on the six measures of well-being (20-item scales inthe first study and 14-item scales in the second) in relation to positiveand negative affect, affect balance (Brad burn, 196 9), the Life Satisfac-tion Index (Neugarten, Havighurst, & Tobin, 1961), the ZungDepres-sion Scale (Zung, 1965), the Center for Epidemiologic Study Depres-sion Scale (CE S-D; Radloff, 1977), and single-item indicators of overallhappiness and life satisfaction. (Se e published studies for details.)

ResultsItem and Scale Analyses

Preliminary analysis (results not shown) indicate that the 18items continue to meet psychometric criteria, with each itemcorrelating strongly and positively with only its own scale.2 Scaleintercorrelations are modest (see Table 1), ranging from .13(e.g., Purpose in Life and Autonomy) to .46 (Self-Acceptanceand Env ironmental M astery ). Estimates of internal consistency(alpha)3 coefficients were low to modest, ranging from .33(Purpose in Life) to .56 (Positive Relations With Others). Thealpha coefficient is a conservative estimate of internal reliabilityfor most (congeneric indicators) scales (Bollen, 1989). Themodest alpha coefficients likely reflect the small number of in-dicators per scale and the fact that items were chosen to repre-sent the conceptual breadth within each construct (see ratio-nale for item selection in Measures and Procedure section)rather than to maximize internal consistency.

Age and Sex Differences in Weil-BeingMean-level analyses, despite dram atic red uctions in depth ofmeasurement, replicated many prior findings (see Figure 1).

The subscales of Purpose in Life, F(2, 921) = 19.8, p < .001;and Personal Growth, F ( 2,921) = 16.4, p < .001; continued toshow decremental age profiles (with scores of the oldest respon-dents significantly lower than those of the two younger agegroups4); Environmental Mastery, F(2, 921) = 3.05, p < .05,also continued to show age increments (with both older groupsscoring significantly higher than young adults), and self-accep-tance showed no age differences. Autonomy again showed ageincrements, F(2, 921) = 4.97, p < .01, but only from youngadulthood to midlife. Although the two previous studies indi-cated no age differences in Positive Relations, these data showedincremen tal scores with age, F( 2,921) = 7.12, p < .001 (olderrespondents scored higher than both younger age groups). Theonly scale that showed significant sex differences was PositiveRelations With Others, F( 1,921) = 8.94, p < .01, with womenagain scoring higher than men .Structural Analyses

We estimated the parameters and fit of the measurementmodels with LISREL 7.2 (Joreskog & Sorbom, 1989). In allmodels the measurement error was assumed to occur ran-domly our goal was not to improve thefitof the measurementmodel but to determine whether the "clean" model (i.e., nocorrelated measurement e rror)fithe data w ell. Besides the the-oretical model, several other models capturing distinct and per-haps more parsimonious explanations for the structure of thedata also were assessed. For example, we estimated a unidimen-sional m odel in w hich all indicators loaded on a single factor ofwell-being. We also explored artifacts of mea surement by esti-

2 Nearly all of the item-to-scales coefficients are statistically signifi-cant, a reflection of the large sample size and, hence, statistical power.3 To be consistent with other analyses in the a rticle, we used listwisedeletion of missing data when comp uting alpha coefficients. The samplesize after listwise deletion is n = 928, compared with a total sample sizeof 1,108.4 Contrasts between each age group were made with Tukey's HonestlySignificant Differences procedu re, with alpha set at .05.


4/9

722 CAROL D. RYFF AND COREY LEE M. KEYES

Young Midlife OlderDimensions of Well-Being

I-Self Acceptance O Positive Relations A Personal Growth A Purpose in Life #E n v . Mastery O Autonomy

Figure 1. Age differences on the six 3-item measures of psychological w ell-being. Env. Mastery = E nvi-ronmental Mastery.

mating two types of models, each suggesting tha t respondentsanswer questions to portray a positive self-image. The firstarti-factual model extended the argument that a single dimension(i.e., well-being) explained the stru cture of the data and furtherspecified that negatively worded items loaded on an artifact di-mension. A second artifactual model suggests that people en-gage in an agreement-disagreement bias to portray a positiveself-image: They agree with all positively worded, and disagreewith all negatively worded, item s. Thus, we estimated a m odelwith two latent artifact constructs, one for positively wordeditems and a second for negatively worded items. Next, we com-pared the fit of the theoretical m odel with the single factor andartifactual models with a super-factor model in which the sixlatent constructs are effects of a second-order latent constructcalled psychological well-being.Descriptive analysis indicates that the marginal distributionsof nearly all items are skewed substantially (results not show n).We therefore used weighted least squares estimation, whichproduces distribution-free and asymptotically unbiased andefficient estimates. In theory, weighted least squares is superiorto maxim um likelihood estimation forfittingmodels with indi-cators whose distributions are non-normal and skewed, espe-cially as sample size increases (Bollen, 1989; Boomsma, 1983;Joreskog & Sorbom, 1989) .5 We produced the variance-covar-iance matrix and asym ptotic covariance ma trix of the estimatedvariances and covariances using PRE LIS (Joreskog & Sorbom,1988). Using listwise deletion of missing data, we purged 180respondents from the sam ple. T he conclusions using mean-sub-stitution of missing data were unchanged. We present only theresults using listwise deletion of missing data.Table 2 displays three indices of the overall fit of each m odel.

Small and statistically nonsignificant values of the chi-squarestatistic indicate good-fitting models. However, none of themodelsfithe data according to chi-square, which is sensitive tosample size and sample variability (Bollen, 1989; Raykov,Tomer, & Nesselroade, 1991). We therefore report the Bayesianinformation criterion (BIC; Raftery, 1993), which is increas-ingly likely to indicate good fitting models as sample size in-creases. Positive values of BIC indicate poor-fitting models, andnegative values indicate good-fitting models. Moreover, smaller(i.e., increasingly negative) values of BIC indicate increasinglybetter-fitting models. Last, the adjusted goodness-of-fit index(AGFI) also indicates the overall fit of the model. The AGFIvaries between 0.0 and 1; values that approach unity indicatebetter models. As a rule of thu mb , values of AGFI .90 or higherindicate very-good-fitting models.Com paring th e overall fit indices for the first four models inTable 3 reveals that the six-factor model is the best-fittingmodel. The ch i-square value is lower, AGFI is higher, and BIC isincreasingly negative. The six-factor model, in fact, reveals avast improvement in fit over each of the preceding m odels, par-ticularly the single-factor model. In t urn , Model 5the super-factor m odelproposes that th e six factors are a function of, orare caused by, another latent construct. That is, each of the sixfactors belongs to a single conceptual domain called well-being.

3 There are trade-offs to each estimation technique. O ne must be con-cerned with the robustness of maximum likelihood estimates and theconsistency of asymptotic, distribution-free estimates. The conclusionsof this study, however, remain the sam e when either type of estimator isused (results of maximum likelihood estimation are available fromCarol D. Ryff on request).


5/9

STRUCTURE OF WELL-BEING 723Table 2Confirmatory Factor Analyses: Indices ofFit Based on Weighted Least Squares Estimation

Model df AGFI BIC1. Single-factor2. Single-factor and

negative item artifact"3. Two-factor, negative andpositive item artifact4. Six-factor5. Second-order, singlesuper-factor

531.9442.6488.2339.1378.7

135127134120129

.85

.86

.86.89

.89

-38.18-93.70-47.70-167 . 64

-166 .04Note, n = 928. AGFI = adjusted goodness-of-fit index; BIC = Bayesian information criterion. The correlation between the latent constructs, overall well-being and the negative item artifact, is con-strained to be 0.

This model is more restrictive and parsimonious than the six-factor model. Comparing the BIC statistic, the super-factormodel fits the data better than the six-factor model. Accordingto Raftery (19 93 ), the BIC statistic should increase at least 10points for each degree of freedom lost to conclude confidentlythat the more complex model (i.e., the model with fewer degreesof freedomM odel 4 ) fits the data better than the m ore parsi-monious m odel (i.e., Model 5 ). Th e difference in the BIC sta-tistic is only about 2 with a difference of 9 degrees of freedom.It is important to emphasize that the super-factor model isnot the same as the single-factor model. The super-factor modelsays that six factors fit the data and that these six factors mea-sure a single latent construct called psychological well-being.Parsimony exists, but at a higher order. Thus , there is a hierar-chical structure in which general well-being has its effectsthrough the six content domains specified a priori by guidingpsychological theory. The age analyses further clarify that lifecourse effects on well-being cannot be explained with a generalfactor because these age effects are n ot uniform different di-mensions of well-being show different age profiles.Table 3 contains the estimates of the co rrelations between theunmeasured, latent constructs. The estimated correlations be-tween the latent construct are im porta nt for descriptive reasonsand, in this case, to highlight the necessity of theory-drivenstructural analysis. Most of the correlations are modest, aroun d.50. However, the correlation between Self-Acceptance and En-

Table 3Correlations Among the Latent Constructs, From WeightedLeast Squares Estimates of he Six-Factor Model

Construct1. Self-Acceptance2. Positive Relations3. Environmental Mastery4. Personal Growth5. Autonomy6. Purpose in Life

1.65.85.53.53.55

2

.65.31.24.30

3

.56.59.38

4

.51.64

5

.39

6

Note, n = 928. All t valuesthat is, the ratio of the estimated coeffi-cient to its estimated asymptotic standard errorare statistically sig-nificant at the .01 level.

vironmental Mastery is very high. Such a high correlation be-tween latent constructs could indicate redundancy or sharedsources of variance. We note, however, that despite the apparentoverlap, Environmental Mastery and Self-Acceptance exhibitdifferent age profiles. Th us, possible redun dancy in structuralanalyses is contrasted with apparen t distinctness in life courseanalyses, illustrating the importance of theory and multiplemodes of testing it in assessing the structure of the well-beingdomain.Correlations With Other Measures

Table 4 provides zero-order correlations from three separatestudies of each of th e six scales of well-being with other prom i-nent indicators of well-being, namely, happiness, life satisfac-tion, and depression. Across these data sets, measures of happi-ness (affect balance or single-item ind icators) show modest tostrong associations with Self-Acceptance and EnvironmentalMastery, somewhat weaker links with Purpose in Life, and stillweaker ties to Positive Relations W ith Others, Personal Grow th,and Autonomy. Parallel, although generally stronger, patterns ofassociation a re evident for life satisfaction (measure d both as amulti-item scale and as single-item indicators)strongest re-lations are evident for self-acceptance a nd env ironmental mas-tery, with the remaining coefficients showing weak to modestassociations. Finally, the multiple indicators of depression showconsistently negative associations with all dimensions of well-being, with the strongest patterns again evident for Self-Accep-tance and Environmental Mastery.

DiscussionThe purpose of the present study was to test the proposedtheoretical structure of a multidimensional model of psycho-logical well-being. Drawn from points of convergence in priortheories of life course development, clinical ac coun ts of positivefunctioning, an d mental h ealth conceptions, the model includessix distinct components of psychological wellness: Self-Accep-tance, Environmental Mastery, Purpose in Life, Positive Re-lations With Others, Personal Growth, and Autonomy. Con-firma tory factor analyses with da ta from a nationally represen-tative sample supported the proposed multidimensional


6/9

724 CAROL D. RYFF AND COREY LEE M. KEYESTable 4Correlations Between Theory-Based Scales oj Well-Being and Prior Measures

StudyRyff( 1989b)*

Ryffet al.(1994)b

Present study

Prior m easures1. Happinessa. Affect balanceb. Negative affectc. Positive affect2. Satisfactiona. Life Satisfaction Index3. Depressiona. Zung DepressionScale1. Happinessa. Single item, global2. Satisfactiona. Single item, global3. Depressiona. CES-D1. Happinessa. Single item, amou ntduring past month2. Satisfactiona. Single item, rate lifeoverall3. Depressiona. Dysfunctional energyb. Dysfunctional affect

SA

.55- . 41.41.7 3

- .59

.5 4

.6 4- . 70

.36

.42- . 32- .45

PR

.30- .19.2 6

.4 3

- . 33

.38

.40- .46

.2 6

.3 5- . 22- .35

New scalesPL

.42- .29.45

.59

- . 6 0

.41

.5 5- .56

.1 3

.10- . 0 5 - . 14

PG

.25- . l l c.3 6

.38

- . 48

.16

.21- .22

.15

.18- .18- .17

A U

.36- .30.26.26

- .38

.31

.3 0- .48

.0 8

.12- . 14- .18

EM

.62- .51.42.6 1

- . 6 0

.51

.6 1- . 6 8

.40

.39- .41- .50

Note. SA = Self-Acceptance; PR = Positive Relations With Others; PL = Purpose in Life; PG = PersonalGrowth; AU = Autonomy; EM = Environmental Mastery; CES-D = Center for Epidemiological StudyDepression Scale (Radloff, 1977)."Affect Balance (Bradburn, 1969), Negative Affect, and Positive Affect are the subscales that compriseAffect Balance; Life Satisfaction Index (Neugarten et al., 1961); Zung Depression Scale (Zung, 1965).b The h appiness item is: "All things considered, how happy are yo u?" The satisfaction item is: "Thinkingabout your life as a whole, how satisfied are you? "0 Not statistically significant at the .05 alpha level. All other coefficients are statistically significant at leastat the .05 alpha level.

structure: The model that best fit the data was one of six pri-mary factors joined together by a single higher order factor. Thismodel showed d rama tic im provement in fit over suggested al-ternatives, especially the single-factor model. The theoreticalformulation of well-being was thus suppo rted as a m ultifaceteddomain encompassing positive self-regard, mastery of the sur-rounding environm ent, quality relations with others, continuedgrowth and development, purposeful living, and the capacity forself-determination.The data also point to the replicative consistency of age andsex differences on these various aspects of well-being. Acrossmultiple investigations having wide variation in depth of mea-surement (20-item, 14-item, and 3-item scales), declining ageprofiles were obtained on Purpose in Life and Personal Growth,incremental scores were evident for Environmental Masteryand Autonomy, and no age differences were obtained for Self-Acceptance. Patterns for Positive Relations varied betweenshowing no age differences or increm ental patterns. Longitudi-nal data are obviously needed to clarify whether these age pro-files represent maturational changes, or cohort differences.Whatever the "source" of these differences, the results un-

derscore the diversity of life course and cohort profiles of well-being. Finally, across all studies, women were found to scorehigher than men on Positive Relations With Others. Depa rtingfrom prior results including cross-cultural comparison (Ryffetal., 1993), these data did n ot show that women also had higherscores on personal growth. Presumably, it is the extensive reduc-tion in content of the growth scale, or the difference in sampling,that accounts for this divergence from prior studies.Com parison of the theory-based indicators of well-being with

other frequently used measures indicated moderate to strongassociations between two scales (Self-Acceptance and Environ-mental M astery) and single- and multi-item scales of happiness,life satisfaction, and depression. However, the remaining fourdimensions of well-being (Positive Relations With O thers, Pur-pose in Life, Personal Growth, Autonomy) showed mixed orweak relationships with these prior indicators. Continued em-pirical reliance on these earlier indices thus translates to neglectof key aspects of positive functioning emphasized in theoreticalaccounts.It is important to recognize that these guiding theories givesurprisingly little com men tary to ha ppiness or positive affect as


7/9

STRUCTURE OF WELL-BEING 725a defining feature of human wellness. In fact, it has been arguedthat certain aspects of positive functioning, such as the realiza-tion of one's goals and purposes, require effort and disciplinethat may well be at odds with short-term happiness (W aterman,1984). In addition, philosophical accounts cau tion against a fo-cus on happiness as the ultimate good in life. History providescountless examples of those w ho lived ugly, unjust, or po intlesslives who were nonetheless happy (see Becker, 1992). Evenwhen present in exemplary lives, philosophers construe happi-ness not as an end in itself but a b yproduc t of other, more noblepursuits (Mill, 1873/1989). These observations, combinedwith the finding that most people (even the disabled, abused,and unemployed people) report themselves to be happy(Diener, 1993; Taylor & Brown, 1988), raise questions(theoretical, philosophical, empirical) about the scientific at-tention lavished on happiness and positive affect, particularly atthe expense of other aspects of positive functioning. Com pre-hensive accounts of psychological well-being need a lso to pro bepeople's sense of whether their lives have purpose, whether theyare realizing their given potential, what is the quality of theirties to o thers, and if they feel in charge of thei r own lives.Apart from expanding the substantive meaning of psycholog-ical well-being, our call to reexa mine the co ntours of positivefunctioning illustrates the complexity involved in defining andassessing structure within a particular domain. Ideally, struc-tural analyses begin with a well-articulated theoretical frame-work, which is then operationalized and tested with procedu resdesigned to assess the proposed theory, such as confirmatoryfactor analysis. However, in the process, it becomes apparentthat final answers do not end with examination of factorialstructure . In this investigation such analyses indicated tha t twoof the six theoretical con structs (Self-Acceptance and Environ-mental Mastery) were highly correlated. In other w ords, datanot theorysuggested a possible five-factor model, whichwould combine indicators of Self-Acceptance and Environm en-tal Mastery. Examination of structure through other analyses,namely, life course profiles, showed, however, that Self-Accep-tance and Environmental Mastery exhibited distinct age pro-files (the former showing little variation by age, the latter show-ing incremental age differences). Analyses of additional groupdifferences (e.g., by social class, ethnicity, or cu ltu re) would fur-ther inform understanding of the basic structure of well-being.Still other pertinent information pertains to how the differentdimensions of positive functioning vary as outcome measuresin process-oriented studies (see Heidrich & Ryff, 1993a, 1993b;Ryff & Essex, 1992; Ryff et al., 1994; Schmutte & Ryff, 1994;Tweed & Ryff, 1991; Van Riper, Ryff, & Pridham, 1992). To theextent that life transitions affect some, but not other dimensionsof wellness, further evidence is garnered for the multidimen-sionality of the w ell-being domain. Finally, assessments of struc -ture need to incorporate method variation. All of the analysesdescribed-thus far rely on self-report techniques, which may in-clude self-presentation biases. Augmenting these data with ob-servational methods, or data obtained from significant others(e.g., spouses), would strengthen the veracity of the hypothe-sized model. In sum, mapping the fundamental structure ofpsychological well-being is a multitask agenda, requiring ongo-ing syntheses of diverse sources of evidence. For now, we offer

the provisional conclusion tha t there is mo re to being well thanfeeling happy and satisfied with life.References

Allport, G. W. (1961). Pattern an d growth in personality. New York:Holt, Rinehart & Winston.Andrews, F. M. (1991). Stability and change in levels and structure ofsubjective well-being. Social Indicators Research, 25, 1-30.Andrews, F. M., & McKennell, A. C. (19 80) . Measures of self-reportedwell-being. Social Indicators Research, 8, 127-156.Andrews, F. M., & Withey, S. B. (1976). Social indicators of well-being:America's perception of ife quality. New York: Plenum.Becker, L. C. (199 2). G ood lives: Prolegomena. Social Philosophy andPolicy, 9(2), 15-37.Birren, J. E., & Renner, V. J. (1980). Concepts and issues of mentalhealth and aging. In J. E. B irren & R. B. Sloane (Eds.), Handbook ofmental health an d aging (pp. 3- 33 ). Englewood Cliffs, N J: PrenticeHall.Bollen, K. A. (198 9). Structural equations with latent variables. New

York: Wiley.Boomsma, A. (1983). On the robustness ofLISREL (maximum likeli-hood estimation) against small sample size and non-normality. Am -sterdam: Sociometric Research Foundation.Bradburn, N. M. (1969). The structure of psychological well-being.Chicago: Aldine.Bryant, F. B., & Veroff, J. (1982). The structure of psychological well-being: A sociohistorical a nalysis. Journal of Personality an d SocialPsychology, 43 , 653-673. Buhler, C. (19 35 ). The curve of life as studied in biographies. Journalof Applied Psychology, 19, 405-409.Campb ell, A., Converse, P. E., & Rodgers, W. L. (1976). The quality ofAmerican life: Perceptions, evaluations, an d satisfactions. New \brk:Russell Sage Fo undation.Diener, E. (1984). Subjective well-being. Psychological Bulletin, 95,542-575.Diener, E. (1993). Most Americans are happy. Unpublished manu-script, Depa rtmen t of Psychology, University of Illinois.Diener, E., & Larsen, R. J. (1993). The experience of emotional well-being. In M. Lewis & J. M. Haviland (Eds.), Handbook of emotions(pp. 405-415). New York: Guilford Press.Diener, E., L arsen, R. J., Levine, S., & Emmons, R. A. (1 985). Intensityand frequency: Dimensions underlying positive and negative affect.Journal ofPersonality and Social Psychology, 48, 1253-1265.Diener, E., Sandvik, E., & Pavot, W. (19 91) . Happiness is the frequency,not the intensity, of positive versus negative affect. In F. Strack, M.Argyle, & N. Schwartz (Eds.), Subjective well-being: An interdisci-plinary perspective (pp. 119-139). Elmsford, NY: Pergamon Press.Erikson, E. (1959). Identity and the life cycle. Psychological Issues, 1,18-164.Green, D. P., Goldman, S. L., & Salovey, P. (199 3). Measurement errormasks bipolarity in affect ratings. Journal of Personality an d SocialPsychology, 64 , 1029-1041.Groves, R. (198 9). Survey errors an d survey costs. New York: Wiley.Headey, B., Kelley, J., & Wearing, A. (1993). Dimensions of mentalhealth: Life satisfaction, positive affect, anxiety and depression. So -cial Indicators Research, 29, 63-82.Heidrich, S. M., & Ryff, C. D. (1 993a ). The role of social comparisonprocesses in the psychological ad aptation of elderly a dults. Journal ofGerontology, 48, 127-136.Heidrich, S. M., & Ryff, C. D. (19 93b). Physical and mental health inlater life: The self-system as m ediator. Psychology an d Aging, 8, 327-338.Jahoda, M. (1958). Current concepts of positive mental health. NewYork: Basic Books.


8/9

726 CAROL D. RYFF AND COREY LEE M. KEYESJoreskog, K. G.,& Sorbom, D. (1988 ). PRELIS:A program for multi-variate data screening and data summarization (2nd ed.). M oores-ville, IN: Scientific Software.Joreskog, K. G., & Sorbom, D. (1989). LISREL 7:A guide to the pro-gram and applications (2nd ed.). Chicago: SPSS.Jung, C. G. (1933). Modern man in search of a soul(W. S. Dell & C. F.Baynes, T rans.). New York: Hartcou rt, Brace & World.Maslow, A. H. (1968). Toward a psychology of being (2nd ed.). NewYork: Van Nostrand.Mill, J. S. (1989). Autobiography. London: Penguin Books. (Originalwork published 1873)Neugarten, B. L. (197 3). Personality change in late life: A developmen-tal perspective. InC. Eisdorfer & M . P. Lawton (Ed s.), Th e psychol-ogy ofadult development and aging (pp.3 11-3 35). Washington, DC:American Psychological Association.Neugarten, B. L., Havighurst, R., & Tobin, S. (1961). The measure-ment of life satisfaction. Journal ofGerontology, 16, 134-143.RadlofF, L. (1 977 ). The CE S-D S cale: A self-report depression scale forresearch in the general population. Applied Psychological Measure-

ment, 7,385-401.Raftery, A. E. (1 993 ). Bayesian m odel selection instructural equationsmodels. In K. A. Bollen & J. S. Long (Eds.), Testing structural equa-tion models (pp. 16 3-180). Newbury Park, CA: Sage.Raykov, T, Tomer, A., & Nesselroade, J. R. (1991). Reporting struc-tural equation m odeling results in Psychology an d Aging: Some pro-posed guidelines. Psychology and Aging, 6, 499-503.Rogers, C. R. (1961). On becoming a person. Boston: HoughtonMifflin.Ryff, C. D. (1985). Adult personality development and the motivationfor personal growth. In D. Kleiber & M. Maehr (Eds.), Advances inmotivation an d achievement: Vol. 4. Motivation an d adulthood (pp.55-92). Greenwich, CT: JAI Press.RyfF, C. D. (1989 a). Beyond Ponce de Leon and life satisfaction: Newdirections inquest of successful aging. International Journal of Be-havioral Development, 12 , 35-55.Ryff, C. D . (1989b ). Happiness is everything, or is it? Explorations on

the meaning of psychological well-being. Journal of Personality andSocial Psychology, 57, 1069-1081.

Ryff, C. D. (19 91 ). Possible selves in adulthood and old age: A tale ofshifting horizons. Psychology and Aging, 6, 286-295.Ryff, C. D. (19 95 ). Psychological well-being in adult life. Current Di-rections in Psychological Science, 4, 99-104.Ryff, C. D., & Essex, M. J . (19 92 ). The interpretation of life experienceand w ell-being: The sample case of relocation. Psychology and Aging,7,507-517.Ryff, C. D., Lee, Y. H., Essex, M. J., & Schmutte, P. S. (1994). Mychildren and me: Midlife evaluations of grown ch ildren an d self. Psy-chology and Aging, 9, 195-205.Ryff, C. D., Lee, Y. H., & Na, K. C. (1993, November). Through thelens of culture: Psychological well-being at midlife. P aper presentedat the m eetings of Gerontological Society of Am erica, New Orleans,LA.Sauer, W. J., & Warland, R. (1982). Morale and life satisfaction. InD. J. Mangen & W. A. Peterson (Eds.), Research instruments in socialgerontology: Vol. 1, Clinical and social psychology (pp. 195-240).Minneapolis: University of Minn esota P ress.Schmutte, P. S., & Ryff, C. D. (1994). Success, social comparison, andself-assessment: Pa rents' m idlife e valuations of sons, daughters, andself. Journal of Adult Development, 1, 109-126.Taylor, S. E., & Brown, J. D. (1988). Illusion and well-being: A social

psychological perspective on mental health. Psychological Bulletin,103, 193-210.Tweed, S., & Ryff, C. D. (199 1). Adult children of alcoholics: Profilesof wellness and distress. Journal of Studies on Alcohol, 52 , 133-141.Van Riper, M., Ryff, C. D., & Pridham, K. (199 2). Parental and familywell-being in families of children w ith Down sy ndrome: A compara-tive study. Research in Nursing and Health, 15, 227-235.Waterman, A. S. (1984). The psychology of individualism. New York:Praeger.Waterman, A. S. (199 3). Two conceptions of happiness: Contrasts ofpersonal expressiveness (eudaimonia) and hedonic enjoyment.Journal of Personality and Social Psychology, 64 , 678-691.Watson, D., Clark, L. A., & Tellegen, A. (19 88) . Development and val-idation of brief measures of positive and negative affect: The PANASscales. Journal of Personality and Social Psychology, 54 , 1063-1070.Zung, W. K. (19 6 5). A self-rating depression scale. Archives of GeneralPsychiatry, 12 , 63-70.

(Appendix ollows on next page)


9/9

STRUCTURE OF WELL-BEING 72 7Appendix

Definitions of Theory-Guided Dimensions ofWeil-Being

Self-AcceptanceHigh scorer: possesses a positive attitude toward the self; acknowledges and ac cepts m ultiple aspects of self, includinggood and b ad qu alities; feels positive about past life.Low scorer: feels dissatisfied with self, is disappointed w ith what has occurred in past life, is troubled about certainpersonal qualities, wishes to be different than w hat he or she is.

Positive R elations With OthersHigh scorer: has warm , satisfying, trusting relationships with others; is concerned ab out the welfare of others; capa-ble of strong empathy, affection, and intimacy; und erstands give and take of human relationships.Low scorer: has few close, trusting relationships with others;finds t difficult to be warm, open, and concerned aboutothers; is isolated and frustrated in interpersonal relationships; not willing to make comprom ises to sustain im portantties with others.AutonomyHigh scorer: is self-determining and independent, able to resist social pressures to think and act in certain ways,regulates behavior from within, evaluates self by personal standards.

Low scorer: is concerned about the expectations and evaluations of others, relies on judgm ents of others to makeimpo rtant decisions, conforms to social pressures to think and act in ce rtain ways.Environmental MasteryHigh scorer: has a sense of mastery and competence in managing the environment, controls complex array ofexternal activities, makes effective use of surrounding opportunities, able to choose or create contexts suitable topersonal needs and values.Low scorer, has difficulty managing everyday affairs, feels unable to change or improve surrounding context, isunaware of surrounding opportunities, lacks sense of control over external world.

Purpose in LifeHigh scorer: has goals in life and a sense of directedness, feels there is meaning to present and past life, holds beliefsthat give life purp ose, has aims and objectives for living.Low scorer: lacks a sense of meaning in life; has few goals or aims, lacks sense of direction; does no t see purpose inpast life; has no outlooks or beliefs that give life m eaning.Personal GrowthHigh scorer: has a feeling of continued development, sees self as growing and expanding, is open to new experiences,has sense of realizing his or her potential, sees improvement in self and behavior over time, is changing inways thatreflect m ore self-knowledge and effectiveness.Low scorer: has a sense of personal stagnation, lacks sense of improvement or expansion over time, feels bored an duninterested with life, feels una ble to develop new a ttitudes or behaviors.

Received April 15 ,1994Revision received February 5,1995Accepted Fe bruary 7,1995

(best) 55.the structure of psychological well-being revisited

Documents