Social well-being Corey Lee M Keyes Social Psychology Quarterly; Sun 1998; 61, 2; ProQuest Psychology Journals pe. [21 Reproduced with pe Social Psychology Quarterly 1998, Vol. 61, No 2121-140, Social Well-Being* COREY LEE M. KEYES. Emory University The proposal of five dimensions of social well-being, socal iteration, socal contribution, social coherence, social actualization, and social acceptance, is the- breticaly substantiate. The theoretical structure, contract ‘alti, and the Sia srucural sources ofthe dimencions of cal well-being are investigated in two studies Item aid confirmatory factor analyses in both Studies corroborate the theoretical model of social wel- being. The new scales correlate convergently with measures of anomie, generativty, perceived social constraints, community involvement and neighborkood quality. The new scles correlate discriminanty with measures of dysphoria global well-being, physical health and optimism. Multivariate analyses in both sudies substantiate the claim thet social well-being is an achievement, faciated by educational attainment and age. The state and direction ofthe study of adult fancioning are discussed. ‘What is the nature of a well-lived life? Does positive mental health include social challenges and criteria? Inquiry into the nature of well-being should embrace the division of life into public and private tasks, f distinction that has pervaded social psy- chological theory. “The self, for example, is both a public process and 'a private product (James 1890; ‘Mead 1934). Individual differences in neu- roticism and extraversion (Costa and McCrea 1980), self-awareness (Duval and ‘Wicklund 1972; Fenigstein, Scheier and Buss 1975), self-conception (Greenwald and Pratkanis 1984; Trafimow, Triandis and Goto 1991; Triandis 1989) and esteem (Luhtanen and Crocker 1992) characterize people a5 either attentive to situational or internal exi- This research was supported by the Joba D. and Catherine T. MacArthur Foundation Research Network on Sucesefl Midife Development (MID MAC), whose diector i Dr. Orville Gilbert Brim. Data for Study 1ate taken from a local study of Positive Socal Functioning, supported by MIDMAC: data for Study 2 come fcom MIDMAC's national study. am indebted to Hazel Rose Markus, Alice Rossi, Dov Shmotkin, Richard Shweder, Elizabeth J Thomion for their support and feedback fon Study 1 and an earlier dratt of this paper. The feedback received ftom the SPQ reviewers and co- ‘editor, Dr. Lynn Smith-Lovin, improved the paper teasurably. Specal thanks goto ane Allyn Piiavic tnd Carol D. Ry for their indefatigable feedback ‘and guidance. Direct correspondence to the Department of Sociology, 1355 Pierce Drive, “Taybuton Halt Atianta, GA 30322. gencies and information Role theories and concepts delineate the strains and incongruites between personal and social expectations (Biddle 1986; Heiss 1981) or focus on the ways in which people manage the incongruity between private and public life and images (Goffman 1959; also see Shaw and Costanzo 1982). Finally, as adults age they purportedly encounter tasks that force them to choose to adapt through private resignation or public social involvement (i.c.. generativity) (Erikson 1950; also sce Adler 1979). The pl vate and the public sides of life are (wo potential sources of life's challenges, with possibly distinct consequences for judging a weellved hi Despite the distinctions between public and private life, the lending conceptions of audult functioning portray well-being 3s 2 pri- ‘marily private phenomenon. The clinical tra- dition tends to operationalize well-being through measures of depression, distress, anxiety, or substance abuse (see, eg, Tholts 1992). Well-being therefore isthe absence of negative conditions and feelings the result ‘of adjustment and adaptation to a hazardous world. The psychological tradition opera- Aionalizes well-being as the subjective evalu- ation of life via satisfaction and affect (c., “Andrews and Withey 1976; Bradburn 1969; Campbell 1981; Campbell, Converse, and Rodgers 1976; Diener 1984; Gurin, Veroff, and Feld 1960) or personal functioning (Ryff ra jon ofthe copyright owner. Further reproduction prohibited without permission. [— 12 1989; Ryff and Keyes 1995). According to this view, emotional well-being is an excess of positive over negative feelings; personal psychological functioning is the presence of more positive than negative perceived self- attributes such as personal growth. Although the existing models emphasize private fea- tures of well-being, individuals remain} ‘embedded in social structures and communi ties, and face countless social tasks and chal-, lenges. To understand optimal functioning) and mental health, social scientists also should investigate adults’ social well-being (also see Larson 1992, 1996) The purpose of this study is to substanti- ate and testa social model of well-being that reflects postive socal health. Therefore I dis- ‘uss the social nature of fife and its challenges, because such challenges might be criteria that individuals use to assess the quality of their lives 1 propose operational definitions and indicators of social well-being, In two repre- sentative samples, F examine the theoretical structure and validity of each scale, as well as some of the social structural origins of each dimension of socal well-being. Symptoms of Life's Social Challenges Social health, or at least its absence, is a preeminent concern in classic sociological theory. Despite the importance of anomie ‘and alienation, Durkheim and Marx also dis- cussed several dimensions of positive social health. Among the potential benefits of pub- lic life are social integration and cohesion, a sense of belonging and interdependence, and a sense of shared consciousness and collec: tive fate (Durkheim 1951). The benefits of social life provide a foundation for a global dofinition of a social version of well-being. Social well-being is the appraisal of ‘one's circumstance and functioning in soci- ery. Below I propose and describe several social challenges that constitute possible dimensions of social wellness. Social integration isthe evaluation of the ‘quality of one’s relationship to society and ‘community. Healthy individuals feel that they are a part of society. Integration is therefore the extent to which people feel they have something in common with others who constitute their social reality (eg. their Reproduced with permission of the copyright owner. SOCIAL PSYCHOLOGY QUARTERLY neighborhood), as well as the degree to which they feel that they belong to their communities and society. Social integration draws on conceptions of social cohesion (Durkheim), cultural estrangement and social isolation (Seeman), and class con- sciousness (Marx). In Durkheim's view, social coordination and health reflect indi- viduals’ connections to each other through norms and indicate their fondness for soci- ety, According to Seeman (1959, 1983, 1991; also see Merton 1949), cultural estrange- iment is the cleavage of self from society Estrangement is the rejection of society or the realization that society does not reflect one's own values and lifestyle. Sgcial isola tion is the break ships that provide meaning and support. Like Marv’s coaception of class conscious ness, social integration entails the construal ‘of collective membership and fate Social acceptance isthe construal of soci- ety through the cheracter and qualities of other people as a generalized category. Individuals who illustrate social acceptance trust others think that others are capable of kindness, and believe that people ean be industrious Socially accepting people hold favorable views of human nature (see Wrightsman 1991) and feel comfortable with others (Homey 1945). Social acceptance is the social analogue to personal acceptance: People who feel good about their personali- ties and accept both the good and the bad aspects of their lives exemplify good mental health (Fey 1955; Ryff 1989). Therefore social acceptance of others might be the social counterpart to self-acceptance Social contribution is the evaluation of ‘one’s social valug, It includes the belief that ‘one is a vital member of society, with some- thing of value to give to the world. Social contribution resembles the concepts of effi, and responsibility. Self efficacy isthe Belel Mat ore can petform eatin behav iors (Bandura 1977) and can accomplish spe- cific objectives (Gecas 1989). Social respon- sibility is the designation of personal obliga- tions that ostensibly contribute to society. Social contribution reflects whether, and to what degree, people feel that whatever they doin the worlds valued by society and con- tributes to the commonweal. This construct Further reproduction prohibited without permission.SOCIAL WELL-BEING consistent with Marx's thesis that people are naturally productive (Israel 1971) Alienation is the economic counterpart to the diminution of the perceived value of ‘one’s life and everyday activities. The devel- ‘opmental analogue of social contribution is generative motives and behavior. Midlife, according to Erikson (1950), isa period when ‘adults can act on their desire to contribute to society by molding the next generation into productive members of society (for example, bby mentoring). Social actualization is the evaluation of the potential and the trajectory of society. This is the belief in the evolution of society and the sense that society has potential which is being realized through its institutions and citizens. Healthier people are hopeful about the condition and future of society, and they can recognize society's potential. Socially healthier people can envision that they, and people like them, are potential beneficiaties of social growth. The evaluation of social progress is consistent with Srole’s (1956) interpretation of anomie as the evaluation of, society's potential by the evalustion of the character of society's custodians (for exam- ple, public officials) Similarly, social actual- ization resembles fatalism (Lefcourt 1982) and powerlessness (Seeman 1991). Parallel to self-determination, social ‘actualization is the sonse that society con- trols its destiny. The focus on the realization ‘of social potential also is similar to the theme of potential and its development as self-realization (Maslow 1968), eudaimonic happiness (Waterman 1993), and personal growth (Ryéf 1989), Optimal functioning, in art. is openness to expericnce and the desire and effort 0 grow continually. Social ‘actuallzation captures these ideas of growth and development. Social coherence is the perception of the ‘quality, organization, and operation of the social world, and it includes a concern for knowing about the world. Healthier people not only care about the kind of world in which they live, but also feel that shey.cap understand what is ha ‘round them. Scarpeople do not dele thomseives tht they live in a perfect world; they have main- tained or promoted the desire to make sense of life. Reproduced with permission of the copyright owner. 123 Social coherence is analogous to mean- inglessness in life (Mirowsky and Ross 1989; Seeman 1959, 1991), and involves appraisals that society is discernable, sensible, and pre- dictable. Psychologically, healthier individuals see their personal lives as meaningful and ‘coherent og 1989). The sense of personal coherence, according to Antonovsky (1994), may be a marker of health: individuals who have coherence attempt to maintain coher- cence when faced with unpredictable and trau- matic life events In sum, perspectives rooted in philoso- phy, social psychological theory and cultural analysis argue cogently for emphasizing the social equally with the personal nature of well-being. Themes of integration, social involvement, and public consciousness in classical sociological theory, which are reiter- ated in social psychological conceptions of self in society, suggest the social challenges faced by adults Social Siructural Sources of Social Well-Being People do not begin or maintain the quest for social well-being with the same assets, Social structure should constrain or facilitate individuals’ ability and opportunity to respond successfully to the social che lenges of life. Social stratification and aging ae prominent structural aspects of life and hhave been the subject of prior research on the sources of other dimensions of health and well-being. In particular, eduestional attain- ment and processes of aging generally affect, instrumental resources and self-conceptions. Educational attainment launches young, adults into specific occupations, affecting what they cam and the conditions and peo- ple they encounter daily. Indirectly, through monetary sequelae, education determines the quality of one’s housing and neighbor- hood (Karabel and Halsey 197; Kohn 1969; Kohn and Schooler 1982; Sewell and Hauser 1975). Lower socioeconomic status (SES) has been linked consistently to diminished physical and mental health (N. Adler et al 1994), partially because life at lower socioe- ‘conomic levels appears to impair health-pro- moting self-conceptions (Mirowsky and Ross 1989), ‘Whereas higher levels of education Further reproduction prohibited without permission128 should promote social well-being, the rela- tionship of age to social well-being is equiv ocal, Aging and age differences include ‘numerous aspects that are not always consis- tent. For example, despite the apparent con- striction of social positions and activity (Carstensen 1995; Riley, Kahn, and Foner 1994), and the decline in physical health with age, research illustrates how adults seem to age successfully through several mechanisms. (For a discussion of possible self-mechanisms, see Heidrich and Ryff 1996.) Moreover, while some self-concep- tions (such as personal control) apparently decline with age (Mirowsky 1995). other studies find that adults as they age, feel hap- pier and more satisfied with their lives and report higher levels of some dimensions of psychological well-being (Heidrich and Ryft 1996; Ryff and Keyes 1995). The ability to age successfully suggests that some facets of ing, like psychological well: being, may inerease with age. In the two probability samples I investi gate the hypothesis of the social nature of well-being and the hypothesis of the social structural sources of variation in social wel being, and subject these hypotheses to the standard of replication. | operationalize the dimensions ina local probability sample and replicate them in a larger, nationally repre~ sentative sample of adults. Confirmatory factor analyses permit assessment of the utility of the hypothesized five-factor model ‘of social well-being. I then correlate the new scales with extant indicators and scales of constructs and processes that reflect individ uals" perceptions of positive social function- ing. In the single exception I correlate the new scales with a scale of anomie, which measures the absence of social health. Insofar as all of the new scales measure social wellness, they should correlate nega- tively with appraisals of anomie. In particular, social actualization and social acceptance should correlate strongly with anomie because the content of the Srole anomie scale focuses on the perceived quality of society (as does social actualiza- tion) and on the character of other people like public officials (as does social accep- tance). Social contribution, on the other hhand, should correlate strongly with a scale Reproduced with permission of the copyright owner. SOCIAL PSYCHOLOGY QUARTERLY (ez, generativty) that measures how great- Jy individuals perceive that they are valued by others, are used as social resources (such 43s sources of advice), and contribute to oth- ers! well-being People who feel socially integrated, ‘lose to and deriving comfort from others in their community, should feel that they live in a vital and healthy neighborhood. Socially integrated individuals therefore should per- ceive their neighbors as trustworthy and their neighborhoods as safe. Adults integrat- ed into society also should be likely to vol- unteer to maintain their neighborhood, per- haps because they feel their actions will be valued by others and because they want to maintain the quality oftheir neighborhood. Finally, people who see life as socially coheront should also feel that their private life is coherent. Therefore individuals with higher scores on the social coherence scale should feel that thei personal Hives are more predicable, more controllable, and thus gen- érally sensible and understandable, Im short, insofar as the new scales mea- sure social well-being, socially healthier indi- viduals should not regard society and its eus- todians as unsavory, should perceive them- selves a5 social resources, should care for and feel safe in their communities, and should lead coherent personal lives, Regression analysis of the disparities in social well-being, by education and age, per- mits investigation of the hypothesized struc- tural sources of social well-being. Prior research on the structural sources of psycho- logical aspects of mental health suggests that social well-being should increase as the level of education rises, Aging, however, involves losses and the ability to adapt and age suc- cessfully: Therefore, although social well- being may increase with age, I explore the possibility of a nonlinear relationship between age and social well-being. For example, some dimensions of social well- ‘being could inerease more rapidly because adults face the same social challenges that subside with age or adults adapt with age. However, social well-being could increase less quickly with age because the social chal- lenges of lfe that might intensify cannot be offset by aduls' attempts to age successfully. Further reproduction prohibited without permission,SOCIAL WELL-BEING METHODS Samples Study 1 is a random-digit dialing sample (Waksberg 1978) of adults age 18 or older living in Dane County, Wisconsin. Madison is the seat of Dane County and the capital of the state. All interviews, utilizing computer- assisted technology, were conducted on the telephone by trained persons at the Letters and Science Survey Center at the University ‘of Wisconsin, Madison, After the pretesting and interviewer training, interviews were conducted between April and June 1994, When determining household composition, interviewers selected the household member ‘who had celebrated his or her birthday most recently (see O'Rourke and Bisir 1983). The interview lasted 30 minutes on average. The saruple includes 373 adults, for a response rate of 63%, ‘Study 2is a random-digit-aling sample of noninstitutionalized English-speaking adults age 25 to 74, living in the 48 contigu- ‘ous states, whose household included at least ‘one telephone.! In the first stage of the mul- tistage sampling design, the investigators selected households with equal probability ia telephone numbers, At the second stage, they used disproportionate stratified sam- pling to select respondents, The sample was stratified by age and sex; males between ages 65 and 74 were oversampled. Field procedures were initiated in January 1995 and lasted approximately 13 ‘months. The respondents were contacted by professional personnel; those who agreed t0 participate in the entire study took part in a computerassisted telephone interview lating 30 minutes on average. Respondents then were mailed two questionnaire booklets, requiring about 1.5 hours on average to com- plete. As incentives for participation in the complete study, each respondent was offered $20, a commemorative pen, petiodic reports Of study findings, and a copy of a monograph ‘on the study. The sample consists of 3,032 ‘adults, with a 70% response rate for the tele- phone phase and an 87% response rate for 1 This sample was assembled by MIDMAC, the John D. and Catherine T. MacArthur Foundation Research Network on Successful Midlife Development. 128 the self-aéministered questionnaire phase, or a combined response rate of 61%. Table I reports the demographic charae- teristics of each sample. The data from the national sample data are weighted to adjust for unequal probabilities of household selec: tion and unequal probabilities of respondent selection within households. The sample ‘weight also posstratifies the sample to match the October 1995 Current Population Survey proportions of adults on the basis of gender, age, race, education, and marital status, as well as the proportions of adults living in metro- politan (nonmetro) areas and various regions (northeast, midwest, south and west) of the United States The mean age of the Dane ‘County sample is 43.6 (SD = 15.9), compared with a mean age of 45.3 (SD = 13.5) in the rational sample. In the Dane County sample, fewer respondents are married and more have attained slighily higher levels of educa- tion. Moreover, the Dane County sample is mostly Caucasian, and more households report slightly higher household incomes than inthe national sample. Measures Social well-being. Theoretical concep- tions in Study 1 (Dane County) promoted ‘operational definitions that guided the cre- ation of indicators (Appendix A). On the basis of linguistic clarity and face validity, 1 retained 10 items (ive positive and five neg- ative) per scale and placed them in a ran- dom sequence. After six background ques- tions were asked, the SO items were adminis- tered at the beginning of the telephone interview. Through the unfolding technique (Groves 1989), respondents stated whether they agreed or disagreed, and then judged ‘how much (strongly, moderately, or slightly). ‘The response options therefore range from 1, strongly disagree, to 7, strongly agree. Interviewers were trained to refrain from converting voluntary expressions of uncer- tainty (.e, “don't know" and “neither agree nor disagree") into substantive answers. A total of 87 respondents expressed uncertainty at least once in response to the items retained in Study 1. Expressions of “neither agree nor disagree” are coded to the midpoint of the scale; “don't know” expressions are assigned Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.126 Charatte Madi (Percentages) SOCIAL PSYCHOLOGY QUARTERLY ‘CSpaliscs of Sd Hamlin ts Mico Foasorigns Shei. Stusy Study : ‘Dane County Sample Natignal (Wewhted) Sample ge le SBaissW= 30) i see 3029} sj Aa 3 : Young: tna! ‘sos Midie 388 403 Older : eae 21s? esider : : : 2 ‘ Ma . ye a Ogos Female 388 568 Marital Stat i z ‘Merrie a sie : ‘Nevec rig? ar Oren 1E Lec Caucasian 2 : ‘Atrxan: American 30 © Other : as Education (Upto 22 year, 500 soo lSyear VT pas s 6-year oF mre 34 3 Income pee ls zt ‘314999 oles 4 he 3 $515,000 fo 24908 93 0 ieSM909 7 z 525.00 40 $49.99, yao $80,900 or mate 366 Dane County samp ‘40 S8;older = 6274, ‘Young = 183%, midi the mean of a respondent's existing items on @ scale. In cases in which they agreed or dis agreed but did not report the strength of their feelings, respondents are assigned the mean of the “agree” or “disagree” side of the response scale. Thus, for example, respondents who agreed with an item but did not know how strongly they felt were imputed the mean of the “agree” (strongly, moderately or slightly) side of the response scale. Each scale in Study 2 consists of only three items. (Because of concern about respondents’ fatigue, I did not include all items) In keeping with the self-administered format, all pronouns are first person (in con- trast to second person in Study 1). The response format ranges from 1, strongly dis agree, to 7, strongly agree. Respondents ‘were given a midpoint option labeled “don’t know.” The social well-being items were ‘ordered randomly and embedded in a sec- tion of questions inquiring about social par- Reproduced with permission of the copyright owner. 10-5, older 60-90 National sample: Young » 25 ticipation, social responsibility, and social networks, Validation criteria, Stady 1.1 measured anomie (Srole 1956) with 2 scale consisting of three items used in the General Social Survey (GSS) (Davis and Smith 1994). Respondents indicated whether they agree or disagree that (4) “Most public officials are not really interested in the problems of the average person,” (2) “The lot of the average person is getting worse, not better.” and (3) cis not fae to bring children into the world ‘with the way things Jook for the furure:” The internal (alpha) consistency of the anomie scale is 57, which is identical to the average reliability (Ma. = .S7, SD =.04) of the same scale over six years of the GSS? “Two indicators measure aspects of global psychological well-being, Respondents indi cated whether they felt very somewhat, a lit- 2 GSS data are taken from 1984, 1985, 1987, 1988, 1989, and 1990 Further reproduction prohibited without permission.SOCIAL WELL-BEING 127 tl, to not at all satistied with thei lives right ‘now, On the same scale, respondents were instructed to pretend for 2 moment that they knew their life would end next week, and then to indicate how happy they felt about the life they had been able to live. Finally, respondents indicated whether they had ever ‘worked with other people in their community to solve a problem in their community. Respondents who had been involved then indicated whether their community activity ‘occurred during the past 12 months Validation criteria, Study 2. Respondents indicated whether six indicators of genera tivity described them a lot, somewhat, a lit- tle, or not at all. The internal (alpha) relibil- ity of the modified Loyola Generativity Scale (based on McAdams and de St. Aubin 1992) is 84. Higher scores reveal individuals ‘who feel that they have made contributions to society, have imparted skills or advice to ‘others, like to teach things to others, have hhad a good influence on others, aud feel needed by others. A scale measuring per- ceived neighborhood health consists of four indicators intended to measure feelings of neighborhood trust and safety. Respondents indicated whether the following statements describe their situations a lot, some, a little, cor not at all: (1) “I feel safe being out alone in my neighborhood during the daytime,” (2) “T feel safe being out alone in may neighbor- hood during the night.” (3) “I could call ona neighbor for help if I needed it,” and (4) “People in my neighborhood trust each other." The internal (alpha) reliability of the neighborhood health sale is .66; a higher score indicates a stronger feeling of neigh- ‘bothood trust and safety. Respondents also completed a scale ‘moasuring perceived constraints, which reflects how much individuals perceive obstacles and unpredictable contingencies in their lives (see Lachman and Weaver Forthcoming). Respondents indicated whether they agree or disagree (strongly, somewhat, ora little) with eight self-descrip- tive statements, including the following as examples: (1) “There is little T can do to change the important things in my life." (2) “What happens in my life is often beyond sy control,” (3) “There are many things that interfere with what I want to do.” and (4)“I sometimes feel Iam being pushed around in amy life" The internal (alpha) reliability of the perceived constraint scale is 86. ‘Twelve items measured the experience of dysphoric symptoms. Respondents indicated hhow often, during the past 30 days, they expe- rienced each symptom: “none of the time;"*a litle ofthe time," “some of the time,” “most of the time,” or “all ofthe time.”The symptoms are feeling (1) “so sad nothing could cheer you up.” @2) “nervous” (3) “restless or fid- ety” (4) “hopeless” (5) “that everything was an effort.” (6) “worthless” (7) “cheerful,” (8) “in good spirits” (9) “extremely happy.” (10) “calm and peaceful,” (11) “satisfied.” and (12) “full of life” The internal (alpha) reliability of the dysphoria seale is 92; a higher score indi- cates the experience of more negative and less positive symptoms Finally, during the telephone interview, respondents indicated their physical health as poor, fair, good, very good, or excellent ‘Also duting the telephone interview res dents indicated whether “being optimistic’ described them a lot, somewhat, a litle, or not at all RESULTS In Study 11 retained items with a correct- €d jtem-tosscale correlation of 30 or higher as indicators of their respective constructs Similarly, in Study 2 I retained items (three per scale) that did not suppress the internal reliability estimate as indicators of the latent ‘construct. One indicator of social coherence in Study 2 suppressed the scale internal con- sistency; therefore I omitted it from further analysis. The retained items functioned as indicators of their respective latent constructs in confirmatory factoranalytic models exam- ining the relative explanatory power of the theoretical model of social well-being I used PRELIS 2 (Joreskog and Sorbom 1993a) to create the variance and covariance matrix, ‘and [estimated the parameters and indices of fit of the measurement models using LISREL. £8 (Joreskog and Sorbom 19936). To assess indicator-to-construct overlap in Study 1, I fitted the five-factor theoretical ‘model to the variance-covariance matrix and ‘observed the completely standardized modi- fication index (CSMI). When each indicator Reproduced with permission of the copyright owner. Further reproduction prohibited without permissioneee 128 and latent construct is standardized, the CSMT is the predicted correlation of each indicator with a latent construct other than its own. An indicator with 3 CSMI of 40 or higher is a potentially overlapping item Neat I refitted the theoretical model estimating the effect of the unrelated con- siruct on the potentially overlapping item to the variance-covariance matrix. One item purportedly indicating social actualization overlapped (40, or > completely standard- ined loading) therefore Lomitted it Examination of higher-order sample moments revealed univariate and multivariate skewness and kurtosis in both Study 1 and Study 2. Because of the small sample size in proportion to the number of indicators in Study 1, I identified iter-distribution trans- formations that minimized the value of Mardia's multivariate test of skewness ond kurtosis (see Bollen 1989:423), The conclu- sions were the same regardless of item trans- formations, bt the transformations slighty enhanced the fit of each model. Consequenty, T show the results of maximum likelihood (ML) estimation based on the exponential transformation, which caused the greatest reduction in multivariate skew and kurtosis ‘Because the sample in Study 2 was large enough, I derived parameter estimates through weighted least squares (WLS). Using PRELIS 2 to estimate the asymptotic vari ance-covariance matrix, I then employed WLS to weight the variance-covariance matrix by the inverse ofthe asymptotic vari ance-covariance, As a result, I achieved (in theory) distribution free, asymptotically unbi ased, efficent parameter estimates. In other words, WLS produces more accurate esti- mates than ML as sample size increases, espe cially when the indicator distributions are highly nonnormal (see Bollen 1989; Browne 1984), Because ofthe large sample size, the multivariate skew of the indicator distibu- tions proved highly influential. The fit ofall models was substantially better with WLS. than with MAL. although these estimators pro- duced identical conclusions Therefore I show ‘only the estimate based on WLS. Validation: Seructure ‘Table 2 presents the chi-square index of Reproduced with permission of the copyright owner. SOCIAL PSYCHOLOGY QUARTERLY fit and the descriptive indices of fit of a series cof models fitted to the variance-covariance matrices from Study 1 and Study 2. As a nule of thumb, values of goodness of fit and adjusted goodness of fit indices at .90 or higher signify models that fit the data very closely. The critical N estimates the sample size needed to reject the null hypothesis fora given model at a given level of alpha (05) {sce Bollen 1989). Critical Ms exceeding 200 represent adequate models; increasingly larg er Critical Ns represent even better fitting models. I do not entertain theories about ‘measurement error correlations; that is, mea~ surement error is prestumed to be random. Models 1 through 3 serve as null hypotheses. Model 1 asserts that there are as many latent constructs as indicators (i.c., indicators are not correlated). [n contrast, Model 2 postulates that the relationship among indicators is the result of a single fac- tor, which means that all indicators intercor- relate at about the same level. Model 3, however, posits that the relationship among tual, because it posits that people respond not to the content but to the valence of the scale items, agreeing with all positively phrased items and disagreeing with all nega- tively phrased items ‘Models 4 through 7 represent tests of whether some of the proposed dimensions of social well-being are distinct. Model 4 posits two latent constructs: One causes the indica- {ors of social integration and social contribu. tion, and the other causes the indicators of social coherence, social acceptance and social actualization.} Model 5 posits three latent constructs 10 test the distinctness of social coherence indicators from social acceptance ‘and social actualization indicators. Thus one of these constructs still causes social integra- tion and social contribution, but another causes only social coherence, while the third construct causes social acceptance and social ctualization. Model 6 posits four latent con- structs to test the theory that social integra- 3 The redaced theoretical models do not exhaust ail possible combinations The reduced theoretical ‘models, based on Conceptual similarity, are the best ‘theoretical hunches about what constructs might not be distinct. Further reproduction prohibited without permission.SOCIAL WELL-BEING native and 129 ‘Stacy Maximum Cakelinood? Piel foie F Independence 2. Single Factor 4 Two Factor Aries 4 Two:Factor Reduced Theory 5 Three Factor Reduce Thedry 45, Four Facto: Reuced i Five Factor Theory 2S suady Nah 1: Dane County Sample “Telephone Interview. (We37) Gri (acen) 36.029) 3 8 2399) B29 51478) 88 (31) Self Administered Questionnaire 2 Mosie! ‘Squate "LTndependence: 332 2. Simple Factcr sar |S 3.Two- Factor Aric! a6 Lf TwosFactor Reduced Theory") 1211 3 Three Facior Reshced Theory 11 002 {6 FourFactor Reduced Theory?” 51155740 708 7. Bie Factor Theory Note GFWAGEI) = Govdnes of Gt index (ajurted goodness offisinex):CR’ at GrI(Anrh 7 2887) Te ay 76 6) m8 (38) wed 1) gu Be 9s (92, 6 38 (2) N= 2887) Weighted sas Squares “Analyses in both samples are based on imputation of mean fom a respondent's own extant teams “eApalis based on exponential iansfonnation of tem distro, tion is a latent construct distinct from social contribution. Model 7 investigates whether social acceptance and social actualization are distinct Jatent constructs. In other words, ‘Model 7 isthe proposed five-factortheoreti- cal model of socal well-being. ‘To examine how well a model fits the data, [compared each model with the imme- diately preceding model, using the chi-square contrast. The difference in chi-square values between models is distributed approximately chi-square; degrees of freedom are equal to the difference in the degrees of freedom between models 3m Study 1, all chi-square contrasts show that each successive and more complex model provides a better fit to the data. Ia fact, the theoretical Model 7 is the best-fit- ting model, with a chi-square contrast of 325 (p <.001) to Model 6. The descriptive Reproduced with permission of the copyright owner. indices of fit suggest that the theoretical five-factor model in Study 1 fits the data only adequately. Simulation studies, howev- cr, suggest that the average values of many descriptive indices decrease as sample size decreases and as the number of indicators increases (see Bolen 1989-277), The theoret- ieal model in Study 1 therefore fits well despite the smaller sample size (N = 373) and the large number of indicators (32). Study 2 replicates the superior fit of the theoretical five-factor model. Again, chi- square contrast illustrate that each successive ‘and more complex model fits the data more Closely, In the end, the theoretical five-factor model provides an excellent fit to the data. ‘The chi-square contrast in the theoretical ‘model, in comparison with Model 6, is 875 (p << .005). Moreover, the goodness of fit indices ‘exceed 90 and the critical N is nearly 400, Further reproduction prohibited without permission.130 SOCIAL PSYCHOLOGY QUARTERLY ‘The descriptive statistics of the indicators and the scales also reveal the quality of the theo- retical model Only one estimate ofthe corre- lation of the indicator with its standardized latent construct (ie, validity coefficient) fails to satisfy the 40 rule of thumb (see Appendix B), Whereas the confirmatory factor analyses supported the theory of five distinct latent factors, the descriptive statistics for the observed scales displayed in Table 3 show that all scales intercorrelate positively. Moreover, neatly all scales, even the reduced- item seales in Study 2, exhibit relatively high internal consistency. The scale of social coher- ence exhibits lower internal reliability in Study 1 but relatively higher reliability in Study 2. On the other hand, the scale of social Acceptance was internally consistent in Study 1 but the set of items did not cohere well in Study 2. These lower internal consistencies may be lower because of sampling variation and because I added @ new item to the scale ‘of social acceptance in Study 2 to achieve bal- ‘ance between negative and positive items, Validation: Correlations Table 4 displays the correlations of the new scales with validtion criteria. I use the 2- statistic to assess whether a particular scale of social well-being correlates more strongly, as hypothesized, with a validation criterion. ‘When I employ the average sample size across all correlations in Study 1 (Mn = 360), the standard deviation is .075; therefore the difference between two z-transformed coeffi- cients in Study 1 must be atleast .15 to be sta- tistically significant at the .05 alpha level In ‘Study 2, the standard deviation of the z-statis- tic is .026; the z-ratio therefore must meet or ‘exceed .052 to be statistically significant at the .05 alpha level. In the discussion of results I focus only on whether the hypothe- sized validation correlations are confirmed. “4 ‘Sample size varies for the coitelaions of com- ‘unity action with the scales of social wellness. Particlany imeresting & the correlation af reccot ‘community ating, in which adults who were involved in their communities during the past month (N = 128) ate contrasted with adults who have never been involved (N = 167). As 2 result, the standacd devia tion of the ztatistc is O82, and the zratio must qual or exceed 164 19 be statistically significant at the 5 alpha evel Reproduced with permission of the copyright owner. In Study 1, all new scales correlate nega- tively with anomie, but social actualization and social acceptance correlate more strong Ty. as predicted, with the scale of anomie. As the evaluation of the quality of society, anomie relates to those dimensions of social well-being which indicate the evolution and potential of society and the perception of the character and qualities of generalized others. Feelings of social integration and social con tribution in Study 1 are higher among indi- viduals who have recently been involved in their communities (i. in community action) than among people who have never been involved, As expected, integration is associat- ced with maintaining recent prosocial ties in ‘one’s community. ‘An unanticipated finding was the strength of the association of recent comma: nity action with social contribution. Viewed in retrospect, attempting to solve a problem in one’s community also could be considered productive activity that should be connected with a sense of contribution. Study 2, howev- cr, reveals the distinction between integra- tion and contribution. As expected, social contribution correlates strongly with genera- tivity: People who feel that they are socially ‘valued also feel that they possess the person- al resources and qualities needed to guide ‘others. Social integration, on the other hand, correlates more strongly with the perceived ‘health of one’s neighborhood. Thus people ‘who feel close to others in their community also feel that their neighborhoods are safe and that their neighbors are trustworthy. Finally social coherence correlates strong ly, as predicted, with the scale of pereeived ‘constraints. Individuals who find the social ‘world more unpredictable, more complex, and ‘more incoherent also tond to view their own fives as complex affairs fll of insurmountable ‘obstacles and unpredictable contingencies’ 5 Some new scales in Study 1 are unbalanced. The imbalance, however, does not appear to systematical: ly influence the correlations Social coherence isthe ‘ost imbalanced scale, consisting of only one postive item; the anomie seal, for example, consists of three negatively phased items Sooal atualization, a bal anced scale correlates more strongly, however, than Social coherence with the sale of anomie. Recall (00, Liat the confimatory two-factor artifact model (neg ative and postive) did pootly in explaining he tel Uonship between the items compared with the series of theoretical models of socal well-being - Further reproduction prohibited without permission.SOCIAL WELL-BEING -4 Social Contribution 5 Sadal Acceptance Stuy 1 Dane Coon Me sb Numbecpf tenis °| ‘Alga Reliability uy. 2- Nations Sample Mo SD Number of fess ‘Aiphs Reliability Nove: For all .p 4-01 tule “The scales of social well-being correlate, but do not appear to overlap, with extant measures reflecting psychological concep- tions of mental health, The social well-being scales correlate positively and modestly with the indicators of life satisfaction and life happiness in Stady 1 In Study 2, these scales exhibit consstent- ly modest and positive correlations with the scale of dysphoria. Because of the sample size and statistical efficiency of Study 2, socialite gration correlates slightly stronger with dys phoria than does any other scale of social ‘well-being On the other hand, although dys- phoria correlates relatively strongly with physical health, all scales of social well-being display small but positive correlations with the indicator of subjective physical health. The nw scales of social wellness therefore appear to measure aspects of mental health and life quality related to, ut distinct from, depressive feelings and physical health. None of the social well-being scales cor- relates very strongly with the perception of ‘oneself as optimist (range of f= 17 10.23). ‘This finding suggests that appraisals of sociat well-being are not unduly influenced by ‘optimism or possibly the tendency to exag- gerate positive qualities © Because physical health and optimism are ascer- tained during the inital telephone faterview, the Reproduced with permission of the copyright owner. 3 w 3s Validation: Social Structural Sources ‘Table $ presents the regressions of each scale of social well-being on age and educa tion. Each model estimates a quadratic (and linear) parameter for age to investigate the possible acceleration or deceleration in the predicted increase of social well-being with age. (Only the statistically significant effects shown, however.) All models also adjust for gender, race, and matial status, which could structure educational outcomes and possibly social well-being The remaining background variables (¢.g., income) either follow from ‘education or explain the hypothesized age cif ferences, and therefore are omitted from these models. The intercept represents the predicted level of social well-being for the average 18-year-old (Study 1) or 25-year-old (Study 2)7 ‘one with the social well-being sales may reflect in part, the mode of administration a6 well athe effeetof the time Ing between the tele- phone interview and the sel-administered question- Tin Study 2,1 show the multivariate analyses bated on unweighted data because the results are the same regardless of sample weighting. The compo- ‘ens of the sample weights in Study 2 are not fune- tions ofthe outcomes in this study Therefore multi- variate models based on the unweighted data should therefore be unbiased and efficient ifthe model is ‘specified correctly (soe Winship snd Rail 1994) Further reproduction prohibited without permission.ot Sos Sud Fy Bane County Sapte _ Aomie'Seat. 2) Recent Community Acwan {Nover=D) -Comradnity Action i. Past (Nes gF-0), 2 Global Lite Satsic ‘GlokalHappinessth Lite 1 SapalActualeaton ‘Soeial Acceptance 2S. Sect dmegratign Social Contebution, cial ers Sud National Sampie 3. Gienerativny Scale 2.7 Health af Nejehborhood Seate Perceived Canetruints Seale | Dyephora Spnpton Seale Subjostve Phystal Healy ‘6° Self Described Opumsnc 7.) Social Acustiztion ‘Socin| Coherence Me Sbeo PS : ses Forall 5 <:01 48 Sp. $508, 4 sigpiean 3 pha = ‘The replication of age and educational differences in social well-being strongly sug- gests that social well-being is shaped by prominent social forces. For the average adult, acquisition of education evidently paves the way for successful responses to ‘social challenges of life. With more educa- tion, each dimension of social well-being increases. Thus social wellness, like all other aspects of health (N. Ader et al. 1994), is graded by processes of social stratification. ‘Fach dimension of social well-being also is distributed by age. Four dimensions ‘increase with age. Social coherence, howev- er, decreases with age, pethaps in part because the intelligibility of the world ‘American society is skewed toward the cele- bration of youthful culture. Compared with Reproduced with permission ofthe copyright owner. SOCIAL PSYCHOLOGY QUARTERLY "nfs note, ie hunsred pent -igh sespondonts engaged nly in company acon. 7A engi cnmuny action }o Yh pst ond inever engaged in comm older adults, younger adults may find the world coherent because they can view a ‘world that tends to reflect their own popular ‘culture, Insofar as age has a nonlinear rela- tionship to social well-being, the evidence favors the “deceleration viewpoint”: Although social actualization and contribu- tion increase with age, the increase deceler- ates with each added year of life in both. ‘Study 1 and Study 2. In other words, social ‘well-being is heterogeneous across ages This, finding is consistent with studies showing the diversity of psychological well-being across age groups. For example, older adults. report more environmental mastery but less purpose in life than younger adults (Ryff and Keyes 1995). .. Further reproduction prohibited without permission,SOCIAL WELL-BEING DISCUSSION Life is purportedly divided into private and publie sides, each making its own demands and producing its own conse- ‘quences. Although we recognize the public demands om individuals, social psychologists hhave not always translated social challenges into eriteria that individuals use to evaluate their well-being We have conceived of social health less as the presence of positive crite- ria than as the absence of negative condi- tions such as anomie and alienation Moreover. the emphasis on social structure may skew attention toward the public side of life as a souree of variation in social health rather than inspiring us to explore the social nature of the meaning of health, Ultimately, current research relies on and possibly flects a bias toward psychological concep- tions of well-being. Researchers either focus con clinical symptomatology such as depres- sion oF use global measures of life satisfac tion and happiness. Multidimensional mod- els (Ry 1989) emerge from classic person- ality theories that conceive of the sell as pri- marily private. “The studies reported here present evi- dence giving credibility to the theory that ei social dimensions. 1 rationalized and validated five dimen- sons of social well-being grounded in classic sociological theory and current social psy- chological perspectives. Data from two stud- {es involving cross sections of adults and uti- lizing different modes of administration (Celephone and self-administration) provide strong evidence for the replicability of the latent structure of the new scales and the social structural profiles of dimensions of social well-being. Confirmatory factor analysis showed that the hypothesized five-factor model of social wellness fits the data best in both studies. ‘The findings suggest that putative social challenges are manifold and distinct. ‘The new scales in Study 1 correlate conver gently with anomie and community involve- ment. In Study 2, the new scales correlate convergently with generativity, neighbor- hood health, and perceived constraints. In ‘both studies, these scales correlate with glob- al idicators of lise satisfaction. happiness and dysphoria. Thus the new scales cTeaAy Reproduced with permission of the copyright owner. 133 link, but do not overlap, with extant mea- sures of mental health, which tend to reflect psychological conception of well-being. Bach scale of social well-being, even social actualization, which emphasizes society's potential and progress, correlates only mini- mally with self-defined optimism. In short, this study suggests that life includes numer ‘ous social challenges; therefore well-being includes social dimensions such as coher- cence, integration, actualization, contribution, and acceptance, Society, 100. is a source of variation in social wellness. Social structural profiles cor- roborate the theory that social wellness is an accomplishment. Social well-being increases with education and, in general, with age. Like other measures of mental health and well-being, it is graded by socioeconomic standing, ft remains to be seen, however, whether the same processes (e.g, self-con- ception) and dimensions (e.g., agency) explain how each aspect of social structure affects each type of mental health and well- being. The relatively strong association of prosocial community involvement with aspects of social well-being also supports the posited achievement of a welllived life Individuals involved in their communities during the past 12 months, but not people involved more than 12 months previously, reported feeling more socially integrated and socially contributive than people who hhad never been involved. Despite the unity of measures of social wellness, it is also clear that social well- being like psychological well-being, is relat- 4 10 age in diverse ways. Some aspects of social well-being decrease linearly: others increase linearly, still others increase, but at 1 decelerating rate. The results therefore suggest that the resources, skills, and experi- ence gained through education and its sequelae, s well as through aging processes, are instrumental to negotiating the chal- lenges of social “According to Nisbet (1953:15), and as quoted by Seeman (1959), the various syn- ‘onyms for alienation testify to the impor- tance of the construct of-alienation. The social scientific conception of a well-lived life also relies on the image of the isolated Further reproduction prohibited without permission.Study SOCIAL PSYCHOLOGY QUARTERLY Suwa) 2 4 (v= 368) seas Daae County Sample Bags (Cnn cicd Bam ‘SegibAcusstizton : “AR ae ofS yeursedvcation Pa Aethe move yearnedtucaton 5) 20033 Lain Rae era Soot Aceplance 8 3 “Ame: t me ae 3054 Topeka = 2 ]8 yersedeat, aes oor mote years ection = 338 Alpha Cs a Sodal Imcptmnon, a nae : ons {ipso $2 yeamedacaon! {3 4 — "sto 18 seas eduction Lr mere yers een ‘Alpha Soont Comabation ‘aoe ae ‘Up gsi2 yearsedatio {ied years education Jnormure yeaseducsts ‘spa ‘Sob Cobecens “Ase to Fi year ecto 1015 yets edition {a6 orimoye ysrveducton | 2 Felina fo "Nua Neils Aas, Tower “fats and mas! at individual. That is, the paragons of the good life are happy and satisfied with their private lives, and they possess personal qualities indicative of psychological well-being, We are left to wonder, however, about the quall- ty of such individuals’ lives as lived with and for other people and for society. By opera tionalizing well-being through satisfaction in domains of life such as family and work, sociologists have implicitly questioned the conception of the individual as a disconnect- ed social entity, DUT Have not transformed ‘this notion. Even anomie, the classic concep- tion of the absence of social health, covers Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ‘only the conceptual territory of social well- being that measures society's trajectory and potential (social actuatization) and the char acter of society through people's character {(Gocial acceptance). As Nisbet states life is a ‘quest to belong. The quest for the good life, however, has many meanings, some of which are represented in the proposed five-factor ‘model of social well-being. In addition to exploring life for more social dimensions of well-being, future researchers on positive mental health and ‘well-being face numerous tasks. An immedi- ‘ate task is to investigate the relationshipsSOCIAL WELL-BEING between measures of mental health and well- being that purportedly reflect distinct con- structs. For example, are measures of psycho- logical well-being truly distinct from mea- ‘sures of social well-being? At stake is the hypothesis that the challenges and labor of adulthood originate in the fundamental sion of public and private life. If life, in both its public and its private details, is the source of the dimensions of wellness, to what extent are the social and psychological challenges bound by culture? ‘What, in turn, are the mediators and the var- ious structural sources of each type of well- being? I repeat Thoits’s (1995) call for the tuse of multiple outcome measures in studies of health and well-being. Progress depends, however, on the development and use of models of well-being that measure the array of human functions, ranging not only from negative to positive but also from psycho- logical to social. REFERENCES Adier, Alfred 1979. Superiority and Soctal Inierest. New York: Norton, ‘Adler, Nancy E., Thomas Boyce, Margaret A. ‘Chesney, Sheldon Cohen, Susan Folkman, Robert L. Kabin, and S Leonard Syme. 1994, “Socioeconomic Status and Health.” American Psychologist 841) 15-24 ‘Andrews, Frank M. and Stephen B. Withey. 1976, ‘Soclal Indicators of Well-Being: Americans’ Perceptions of Life Quality. New York: Plenum. Antonovsky, Aaron. 1994, “The Sense of Coherence: An Historical and Future Perspective." Pp. 3-40 in Sense of Coherence ‘and Resiliency: Stress, Coping, and Health, ‘edited by Hl. MeCubbin, E.A. Thompson, ALL Thompson, and JE, Fromer. Madison: University of Wisconsin Press Bandura, Albert, 1977, "Self-Efficacy: Toward 2 Unifying Theory of Behavioral Change.” Peychological Review 84(2):191-215 Biddle, Bruce J. 1986. “Recent Developments in Role Theory.” Annual Review of Sociology 1267-92. Bollen, Kenneth A. 1989. Structural Equations with Latent Variables. New York: Wiley. Bradburn, Norman M. 1969. The Structure of Psychological Well-Being. Chicago: Aldine. Browne, Michael W. 1984. “Asymptotic Distribution Free Methods in Analysis of Covariance Structures” British Journal of Mathematical and Statistical Psychology Reproduced with permission of the copyright owner. 135 3762483 Carstensen, Laura L. 1995. “Evidence for a Life- ‘Span Theory of Socioemotional Selectivity” Current Directions in Psychological Sclence 4151-56. Campbell, Angus. 1981. The Sense of Well Being in America: Recent Patterns and Trends New ‘York: McGraw-Hill ‘Campbell, Angus, Philip A. Converse, and Willard TL. Rodgers’ 1976, The Quality of American Life. New York: Sage. Costa, Paul T. and Robert R. McCrea, 1980. “Lafluence of Extraversion and Neurotieism on Subjective Well-Being: Happy and Unhappy People." Journal of Personality and Socal Psychology 3868-73 Davis James A. and Tom W. Smith. 1994. General ‘Soctal Surveys, 1972-1994: Cumulative Codebook. Chicago: National Opinion esearch Center, University of Chicago. Diener, Ed, 1984. "Subjective Well-Being.” Psychological Bulletin 95(3)'542.75. Durkheim, Emile. 1951. Suicide. New York: Free Press. ‘Duval, Shelley and Robert A. Wicklund, 1972. A Theory of Objective Seif-Awareness. New ‘York: Academic Press Erikson, Erik H, 1950. Childhood and Society Now York: Norton. Fenigsein, Allen, Michael F Scheier, and Arnold 'H, Buss, 1975, “Public and Private Self Consciousness: Assessment and Theory.” Journal of Consulting ond Clinical Peyehology #52221. Fey, WE. 1955. “Acceptance of Others and Its Relation to Acceptance of Self and Other: ‘A Reevaluation” Journal of Abnormal and ‘Social Psychology SO(2)274-76. Geeas, Viktor. 1989. "The Social Psychology of ‘Seif Efficacy” Annual Review of Sociology 15291316, Greenwald, Anthony G. and Anthony R. Pratkanis, 1984, “The Self" Pp. 129-78 in Handbook of Soctal Cognition Vol. 3, edited by RS Wyer and TK. Srull. Hillsdale, Ni: Exlbavm. Goffman, Erving, 1959. The Presentation of Self ‘Everyday Life. New York: Doubleday. Groves, Robert. 1989, Survey Errors and Survey ‘Costs New York: Wiley. Gurin, Gerald, Joseph Veroff, and Sheila Feld.. 1960. Americans View Their Mental Health New York: Basic Books Heidrich, Susan M. and Carol D. Ryf.1996.*The ‘Self in Later Years of Life.” Pp. 73-102 in ‘Aging in the Twenty-First Century: A Developmental Perspective, edited by L. Sperry and H. Prosen, New York: Garland. ‘Heiss, Jerold, 1981. “Social Roles” Pp. 4-129 in “Social Psychology: Sociological Perspectives, Further reproduction prohibited without permission.136 edited by M. Rosenberg and RH. Turner. New York: Basic. Horney, Karen, 1945. Our Inner Conflicis: A Constructive Theory of Neurosis. New York: Norton, Israel, Joachim. 1971. Alienation: From Mars to ‘Modern Sociology. Boston: Allyn and Bacon. James, William. 1890. Principles of Psychology. New York: Holt Joreskog, Karl G. and Dag Sorbom. 19932, PRELIS 2:A Program for ulivariate Data Screening and Data Summorization. ‘Mooresville, IN: Scientific Software, In. 19930, LISREL 8: A Guide to she Program ‘and Applications. Chicago: SPSS, Ine Karabel, Jerome and A.H. Halsey. 1977. Power ‘and Ideology in Education. New York. Oxford University Press. Kohn, Melvin. 1969. Class and Conformity. Homewood, IL: Dorsey. Kohn, Melvin and Carmi Schooler. 1982. “Job ‘Conditions and Personality: A Longitudinal Assessment of Theit Reciprocal Effects” ‘American Journal of Sociology 81-1257-86. Lachman, Margie E. and Suzanne L. Weaver. Forthcoming. “The Sense of Control as Moderator of Social Class Differences in Health and Well-Being.” Journal of Personality and Social Psychology 14): 16- n Larson, James $, 1992. “The Measurement of ‘Social Well-Being.” Social Indicators Research 8285.96, 1996. "The World Health Organization's Definition of Health: Social versus Spiritual Health.” Social Indicators Research 38:181- 2. Lefcourt, Herbert M, 1982. Locus of Control (Current Trends in Theory and Research. 20d ed, Hillsdale, NJ Erlbaum. Luhtanen, Riia and Jennifer Crocker. 1992. "A Self- Collective Self-Esteem Scale: Evaluation of One's Social Iden Personality end Social Psychology Bulletin 18):302-18, Maslow, Abraham H. 1968. Toward a Psychology (of Being. nd ed. New York: Van Nostrand. ‘MeAdams, Dan P. and Ed de St. Aubin, 1992. "A. “Theory of Generativity and Its Assessment through Self-Report, Behavioral Acts, and Narrative Themes in Autobiography.” Journal of Personality and Social Peychology 62(6):1003-15. ‘Mead, George H. 1934. Mind, Self and Society ‘From the Standpoint of a Social Behaviorist (Chicago: University of Chieago Press. ‘Merton, Robert. 1949. Social Theory and Social ‘Structure, Glencoe, L: Free Press. Mirowsky, John, 1995, "Age and the Sense of SOCIAL PSYCHOLOGY QUARTERLY Control.” Social Psychology Quarterly S318, Mirowsky, John and Catherine E. Ross. 1989. Social Causes of Psychological Distress. New York: Aldine. Nisbet, Robert. 1953. The Quest for Community. New York: Oxford University Press. O'Rourke, Diane and Johnny Blair. 1983. “Improving Random Selection in Telephone. Surveys” Journal of Marketing Research 20:408-22, Riley, Mati, R-L. Kahn and A. Foner. 1994, Age ‘and Structural Lag: Society's Failure 1o Provide Meaningful Opporunties in Work, Family, and Leisure. New York: Wiley. Ry/ff, Carol D. 1989, “Happiness Is Everything, or Is It? Explorations on the Meaning of Psychological Well-Being,” Journal of Personality and Social Psychology 546):1069-81, Ryff, Carol D. and Corey Lee M. Keyes. 1995. “The Structure of Psychological Well-Being Rovisited Jounal of Personalty and Social Paychology 64) 19.22. Seeman, Melvin. 1959. “On the Meaning of “Alienation.” American Sociological Review 2785-91. 1983. "Alienation Motifs in Contemporary “Theorizing: The Hidden Continuity of the Classic Themes.” Social Psychology ‘Quarterly 4617184 1991. “Alienation and Anomie." Pp. 291- 374 in Measures of Personality and Social Psychological Attitudes, edited by IP. Robinson, PR. Shaver, and L.S., Wrightsman. San Diego: Academic Press. Sewell, William and Robert Hauser. 1975. Education, Occupation, and Earnings Achievement in the Early Career, New York: ‘Academic Press. Shaw, Marvin E. and Philip R. Costanzo. 1982. ‘Theories of Social Psychology: Ind ed. New York: Mc@raw-Hi Srole, Lea. 1956, “Socal Integration and Certain ‘Corollaries: An Exploratory Study.” American Sociological Review 21:109-16 ‘Thoits, Peggy A. 1992. “Identity Structures and ‘Paychological Well-Being: Gender and Marital Comparisons” Social Psychology Quarerly S50)236-56. 1995, "Stress, Coping, and Social Support Processes: Where Are We? What Next?® Journal of Health and Social Behavior (ena issue) 53-79. ‘Trafimow, David, Harry C Thiandis, and Sharon . ‘Goto, 1991. “Some Tests of the Distinction ‘between the Private Self and the Collective Selt” Journal of Personality and Soctal Paychology 6649-55, ‘Triandis, Harry C. 1989. “The Self and Social Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.SOCIAL WELL-BEING 137 Behavior in Differing, Cultural Contexts” Psychological Review 96(3) 506-20. ‘Waksberg, Joseph, 1978. “Sampling Methods for Random Digit Dialing.” Journal of the American Staistical Association 7340-46 Waterman, Alan S. 1993. “Two Conceptions of Happiness: Contrasts of Personal _Expressiveness (Eudaimonia) and Hedonic Enjoyment.” Journal of Personality and Social Psychology 64678-91 Winship, Christopher and Larry Radbil. 1994 “Sampling Weights and Regression Analysis.” Sociological Methods and Research 23(2)230-57 Wrightsman, Lawzence S. 1991. “Interpersonal Trust and Attitudes toward Homan Nature.” Pp, 373-412 in Measures of Personality and Social Psychological Ausitudes, edited by 1.P. Robinson, PR. Shaver, and LS, Wrightsman. San Diego: Acadeinic Press Corey Lee M. Keyes is an Assistant Professor in the Department of Sociology and the Rollins School of Public Health (Department of Behavioral Sciences and Health Education). He is an ‘associate of the MacArthur Foundation MIDMAC research network. He also is a recent inductee into a research network on “Positive Psychology” organized and led by APA President Martin Seligman and Mihalyi Csikszentmihalyi to build and promote a field of soctal-scientific inquiry into Positive Health. His research explores the social psychology of health and well-being and successful aging. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.