Moderating Effects of Gender and Age on the Relationship Between Self-esteem and Life Satisfaction...

8/6/2019 Moderating Effects of Gender and Age on the Relationship Between Self-esteem and Life Satisfaction in Mainland C

1/10

INTERNATIONAL JOURNAL OF PSYCHOLOGY, 2002, 37(2), 8391

Moderating effects of gender and age on the relationship

between self-esteem and life satisfaction in mainland Chinese

Liwei Zhang Jin-Pang Leung

Beijing University of Physical Education, China The Chinese University of Hong Kong, China

Previous studies indicated that not only individual self-esteem (e.g., Diener & Diener, 1995; Neto, 1993) but also collective self-

esteem (e.g., Crocker, Luhtanen, Blaine, & Broadnax, 1994; Zhang & Leung, 1999) contributed to prediction of life satisfac-

tion. When the effects of individual and collective self-esteems on life satisfaction have been demonstrated, a further question

in life satisfaction research is that whether the relationship between self-esteem and life satisfaction is subject to the inuence

of boundary conditions. The present investigation focused on moderating effects of gender and age on the relationship

between self-esteem and life satisfaction in Chinese people. Participants were 1347 Mainland Chinese (aged from 14 to 88

years, 52.3% female) from three generations. They nished the General Life Satisfaction Scale (Leung & Leung, 1992), Life

Domain Satisfaction Scale (revised from Michalos, 1985), Self-Esteem Scale (Rosenberg, 1965) and Collective Self-Esteem

Scale (Luhtanen & Crocker, 1992). Hierarchical regression analysis indicated that the relationship between collective self-

esteem and general life satisfaction was stronger for the male participants than for the female participants. The effect of indi-

vidual self-esteem on life domain satisfaction was stronger in the male group than in the female group. The effect of individual

self-esteem on life domain satisfaction was stronger in the older people than in the younger people. However the effect of col-

lective self-esteem on life domain satisfaction was stronger in the younger people than in the older people. These results might

reect life task differences and social expectation differences between male and female, younger and older people. Additional

research is needed with other types of samples, especially with samples from some individualist cultures to see whether the

results could generalize to these cultures.

Des tudes pralables ont indiqu que non seulement lauto estime individuelle (p.ex., Diener & Diener, 1995; Neto, 1993)

contribuait prdire la satisfaction dans la vie, mais aussi lauto estime collective (p.ex., Crocker, Luhtanen, Blaine, &

Broadnax, 1994; Zhang & Leung, 1999). Une fois les effets de lauto estime individuelle et collective sur la satisfaction vitale

dmontrs, la nouvelle question qui surgit dans la recherche de la satisfaction vitale est si la relation entre lauto estime et la

satisfaction dans la vie est sujette linuence de conditions limitatives. La prsente tude est centre sur les effets modrateurs

du sexe et de lge sur la relation entre lauto estime et la satisfaction dans la vie chez des chinois. Les participants taient 1347

chinois continentaux (gs de 14 88 ans, 52.3% des femmes) de trois gnrations diffrentes qui ont rpondu la General

Life Satisfaction Scale (Leung & Leung, 1992), la Life Domain Satisfaction Scale (rvise par Michalos, 1985), la Self

Esteem Scale (Rosenberg, 1965) et la Collective Self-esteem Scale (Luhtanen & Crocker, 1992). Lanalyse de la rgression

hirarchique a indiqu que la relation entre lauto estime collective et la satisfaction dans la vie gnrale tait plus forte pour

les participants hommes que pour les femmes. Leffet de lauto estime individuelle dans le domaine de la satisfaction vitale

tait plus fort dans le groupe des hommes que dans celui des femmes. Leffet de lauto estime individuelle dans le domaine de

la satisfaction vitale tait plus fort chez les personnes ges que chez les personnes plus jeunes. Cependant, leffet de lauto

estime collective dans le domaine de la satisfaction tait plus fort chez les jeunes que chez les personnes plus ges. Ces rsultats

retent des diffrences concernant les tches dans la vie et les attentes sociales, ainsi que entre hommes et femmes et entre les

personnes jeunes et les moins jeunes. Des recherches supplmentaires avec dautres types dchantillons sont ncessaires, en

particulier avec des chantillons de cultures individualistes, an de voir si les rsultats peuvent tre gnraliss ces cultures.

Estudios anteriores indicaban que no slo la auto-estima individual (por ejemplo, Diener & Diener, 1995; Neto, 1993), sino

tambin la colectiva (por ejemplo, Crocker, Luhtanen, Blaine, & Broadnax, 1994; Zhang & Leung, 1999) contribuan a la

prediccin de la satisfaccin en la vida. Una vez demostrados los efectos de las auto-estimas individual y colectiva sobre la

satisfaccin vital, la pregunta que surge en la investigacin sobre satisfaccin vital es si la relacin entre la auto-estima y

la satisfaccin vital se sujeta a la inuencia de condiciones limitantes. La presente investigacin se centr sobre los efectos

moderadores del sexo y la edad sobre la relacin entre la auto-estima y la satisfaccin vital en chinos. Los participantes fueron

1347 chinos continentales (entre 14 y 88 aos de edad, 52.3% mujeres) de tres generaciones, quienes contestaron la General

Life Satisfaction Scale (Leung & Leung, 1992), la Life Domain Satisfaction Scale (revisada Michalos, 1985), la Escala de Auto-

Estima (Rosenberg, 1965) y la Collective Self-esteem Scale (Luhtanen & Crocker, 1992). El anlisis de regresin jerrquica

indic que la relacin entre la auto-estima colectiva y la satisfaccin vital general fue ms robusta en el caso de los hombres

Requests for reprints should be addressed to Liwei Zhang, Section of Sport Psychology, Beijing University of Physical Education, 10008

Beijing, PR China (Tel: 86-10-62989565 ; E-mail: [email protected]).

Special thanks should go to the United Board for Christian Higher Education in Hong Kong, which supported this project.


2/10

Psychologists have paid more attention to theoretical andempirical work of peoples subjective wellbeing (SWB)

since the 1960s (Diener, 1984). American Psychologist has

recently even contributed a special issue to the discussion

on positive psychology. The editors (Seligman &

Csikszentmihalyi, 2000) have called on a change in the

psychological research approach from the ill-being model

or the disease model to the wellbeing model or positive

psychology.

What is SWB? Why do people have different levels of

SWB? Who enjoys a higher level of SWB? And how can we

better predict peoples level of SWB? These fundamental

questions have interested and confused human beings for

many centuries because SWB is central to human

existence. The rst question is related to the nature of

SWB, whereas the other three questions are related to

conditions for SWB. The present investigation will try to

address the issue of prediction of SWB.

SWB is a persons evaluative reactions to his or her life

(Diener & Diener, 1995). SWB research is mainly

concerned with questions in two areas. The rst research

area is related to the structure (or components) and

measurement of SWB. Today most researchers agree that

SWB is a hierarchical and multidimensional concept(Diener, 2000; Diener, Suh, Lucas, & Smith, 1999; Feist,

Bodner, Jacobs, Miles, & Tan, 1995). SWB can be

separated into cognitive or affective aspects (Andrews &

Withey, 1976; Diener, 1984; Diener et al., 1999), and state

or trait aspects (see a review by Diener et al., 1999). Life

satisfaction is a cognitive aspect of SWB. It can be divided

further into general life satisfaction and life domain

satisfaction (Diener, 1984; Diener et al., 1999; Veenhoven,

1996). The present investigation will focus on life

satisfaction.

The second research area of SWB is related to the

prediction of SWB and boundary conditions for itsprediction. In the long pursuit of understanding SWB,

psychologists have made great efforts to search for

predictors of life satisfaction and the boundary conditions

for the relationship between life satisfaction and its

predictors. Five kinds of variables have been widely

employed to predict life satisfaction, including

demographic variables, social relationships, personality,

coping, and self-esteem. Self-esteem was suggested to be

one of the strongest predictors of life satisfaction (Diener,

1984). For example, in a study examining relationships

among some psychological variables in institutionalized

elderly residents, Nehrke, Hulicka, and Morganti (1980)found a moderate correlation (r 5 .43) between life

satisfaction and self esteem In a large scale survey

(.55) between these two variables. Campbell, Conversand Rodgers (1976) discovered that self-esteem showed th

highest correlation with life satisfaction of any variab

examined. Neto (1993) found in an adolescent sample th

the strongest predictor of satisfaction with life was sel

esteem, followed by loneliness, which reflects soci

isolation. More recently, in a study involving 131,11

college students in 31 countries, Diener and Diener (199

found a correlation ofr5 .47 between life satisfaction an

self-esteem.

It is obvious that research and theory on self-esteem an

its relationship to life satisfaction has overwhelming

emphasized the individual aspects of the self. That i

research and theory have emphasized how feelings of sel

worth and self-respect are derived from or related to a

individuals personal attributes, competencies, an

standing relative to other individuals. The general publi

especially in the Western culture, appreciates th

importance of individual self-esteem in everyday huma

functioning. Politicians, educators, social workers, an

others have come to hope that individual self-esteem ma

hold the key to understanding and even solving man

social problems and achieving wellbeing.

More recently psychologists began to pay attention tthe differentiation between individual self-esteem an

collective self-esteem. Whereas individual self-estee

refers to feelings and evaluations of self-worthine

(Baumeister, 1998), collective self-esteem refers to feelin

and evaluations of the worthiness of a social group, suc

as racial, ethnic, or work groups, of which one is a memb

(Crocker & Major, 1989). Research evidence shows th

collective self-esteem could also be a potentially stron

predictor of life satisfaction. Crocker, Luhtanen, Blain

and Broadnax (1994) conducted a study on th

contribution of collective self-esteem to psychologic

wellbeing. They found that even after partialing out theffects of individual self-esteem on life satisfaction

collective self-esteem still correlated with certain aspects o

general life satisfaction. Zhang and Leungs study (199

revealed that although individual self-esteem is the be

predictor of athletes general life satisfaction, collectiv

self-esteem is the most powerful predictor of their trainin

and competition satisfaction. These two studies indicate

the limitation of only using individual self-esteem

predict life satisfaction and the necessity of using collectiv

self-esteem as a life satisfaction predictor.

When the effects of individual and collective self-estee

on life satisfaction have been demonstrated, a furthquestion in life satisfaction research is whether th

relationship between self esteem and life satisfaction

84 ZHANG AND LEUNG

que en el de las mujeres. El efecto de la auto-estima individual sobre la satisfaccin en el dominio vital result ms fuerte en

el grupo de los hombres que en el de las mujeres, y en las personas mayores que en el de las ms jvenes. No obstante, el efecto

de la auto-estima colectiva sobre la satisfaccin en el dominio vital fue ms grande en la gente joven que en la de ms edad.

Estos rsultados podran reejar diferencias en las tareas vitales y en las expectativas sociales entre hombres y mujeres, y entre

jvenes y adultos mayores. Se requiere investigacin adicional que emplee otros tipos de muestras, especialmente muestras de

algunas culturas individualistas para determinar si los resultados pueden generalizarse a esas culturas.


3/10

studies indicated that there are some factors which

moderate this relationship. For example, although self-

esteem has been found the best predictor of life

satisfaction among adults (Campbell et al., 1976), parent

relationship has been demonstrated the best predictor of

life satisfaction among adolescents (Leung & Leung,

1992), suggesting that age might be a moderator between

self-esteem and life satisfaction. Another example is that

the relationship between (individual) self-esteem and life

satisfaction has been found stronger in individualist

cultures than in collective cultures (Diener & Diener,

1995), suggesting that culture works as a moderator. These

ndings remind life satisfaction researchers that the link

between self-esteem and life satisfaction is not consistent

when there are differences in peoples background.

Therefore defining the boundary conditions for the

relationship between self-esteem and life satisfaction is

one of challenges that life-satisfaction researchers are

facing. The present investigation will focus on two possible

boundary conditions, i.e., gender and age. The questions

we will try to answer in the present investigation are

whether gender and age are moderators between self-

esteem and life satisfaction.

Past research on life satisfaction largely employed

Western people and similar research employing Mainland

Chinese is lacking. This is surprising because Mainland

Chinese constitute about one-fourth of the worlds

population. Their psychological change and wellbeing

during rapid social, political, and economic development

due to the open-door policy have attracted much attention

from different disciplines of social science, but only in

recent years. Some researchers have begun to study lifesatisfaction in China. For example, in a study on the urban

elderly, Guo (1992) found that among 12 variables,

satisfaction with nance, subjective health, depressive

symptoms, occupation, and close friends became

predictors of life satisfaction in the stepwise regression

model. In a more recent study on the rural and urban

elderly, Xiang, Wu, and Liu (1995) found that in total 17

variables were correlated with life satisfaction in the

stepwise regression model. Among these 17 variables,

subjective health, family harmony, finance situation,

mood, and education contributed most to the explained

variance in life satisfaction.Although these studies helped to improve our

understanding of life satisfaction of Mainland Chinese,

there are at least three reasons for further research. First,

the previous research focused mainly on the elderly (e.g.,

Guo, 1992; Xiang et al., 1995; Xu, 1994; Yang, 1988). Life

satisfaction of adults (no studies found) and adolescents

(only one study found, e.g., Tan, 1998) got much less

attention. However, life satisfaction and the relationship

between life satisfaction and its predictors might differ in

different generations. Second, previous studies in China

neglected the contribution of self-esteem, especially

collective self-esteem, to life satisfaction. However, thecollective aspect of self-esteem should be paid more

attention in a collective culture such as that of China

gender and age without paying attention to possibl

moderating effects of gender and age. These three reason

justify the effort of the present investigation to go furth

into this important area.

Therefore, in our exploration of moderating effects o

the relationship between self-esteem and life satisfaction i

Chinese people, individual self-esteem and collective sel

esteem will be taken as two predictors, general lif

satisfaction and life domain satisfaction will be taken a

two criterion variables, and gender and age will be treate

as two moderators.

METHOD

Participants

Data collection for the present study was conducted in tw

large cities in Mainland China, Beijing and Chongqing

Beijing is located in the northern part and Chongqing i

the southern part of China. Participants came from fouage groups. The rst age group consisted of 427 senio

high school students, the second of 366 universit

students, the third was 360 parents of the senior hig

school students, and the last was 194 grandparents o

those students. The total sample size was 1347. Th

participants ages ranged from 14 to 88 years with a

average of 31.88 (SD 5 19.09). Among them, 46.8% wer

male (631) and 52.3% were female (704). Twelv

participants did not report their gender. Summar

information of sample size, age range, and age mean in ag

and gender groups is presented in Table 1.

Measures

The information about age, gender, and othe

demographic variables was collected on the last page o

the questionnaire booklet.

General life satisfaction

General life satisfaction has been dened as a person

overall evaluation of their quality of life based upon sel

selected standards (Shin & Johnson, 1978). In the presen

investigation, it was measured by the General LifSatisfaction Scale (GLSS), which was originally used i

Leung and Leungs (1992) study. The GLSS is

combination of the Satisfaction With Life Scale (SWLS

Diener, Emmons, Larsen, & Grifn, 1985) and a global lif

satisfaction question. The SWLS consists of 5 item

asking respondents how they feel about their life in variou

ways (e.g., I am satised with my life). The alph

coefcient of SWLS is .87 (Diener et al., 1985). The glob

measure is from a single-item scale by Andrews an

Withey (1976). In the present investigation it was revise

to In general, I feel that my life is terrible so that th

same rating scale could be used for all 6 items. Participantwere asked to check one of seven categories from

(strongly disagree) to 6 (strongly agree) The alph

SELF-ESTEEM AND LIFE SATISFACTION 8


4/10

Life domain satisfaction

Life domain satisfaction was measured by the Life

Domain Satisfaction Scale (LDSS). It was adopted from

Michalos (1985) method but the scope of inquiry was

enlarged from Michalos s 8 domains to 13 domains.

Back-translation was used to improve the linguistic

equivalence of the Chinese version. Participants were

asked to rate their satisfction with 13 life facets on a 7-

point Likert scale (0 5 very unsatisfactory, 6 5 very

satisfactory). The 13 life facets include health, nances,

spouse relationship (if applicable), child relationship (if

applicable), sister or brother relationship (if applicable),

parent relationship (if applicable), peer relationship,

friend relationship, boss (or teacher) relationship, paid

employment or study, housing, recreation activity, and

education. Previous research (Lance, Lautenschlager,

Sloan, & Varca, 1989; Larsen, Diener, & Emmons, 1985)

demonstrated testretest reliabilitie s in the .80s as well asthe construct validity (Andrews & Crandall, 1976; Diener,

1984 ; Scarpell o & Campbell, 1983) of single-item

satisfaction measures such as these. A composite score of

life domain satisfaction was computed by averaging these

13 items. The coefficient alpha for LDSS in this

investigation was .85.

Individual self-esteem

As in the study by Zhang and Leung (1999), individual

self-esteem was measured by Rosenbergs (1965) 10-item

Self Esteem Scale (SES). Its average reliability coefcient is

greater than .80 (Rosenberg, 1965). The Chinese version

came from Ji and Yus (1993) work. An example question

is I feel that Im a person of worth, at least on an equal

plane with others. Items were anchored with the revised

7-point Likert scale of agreement (05 strongly disagree to

6 5 strongly agree). The alpha coefficient in this

investigation was .78.

Collective self-esteem

Collective self-esteem was measured by Luhtanen and

Crockers (1992) Collective Self-esteem Scale (CSES).

This 16-item scale has two forms: general and race-specic. The general form was used in this study with some

revision on the reference groups Originally the

related to gender, race, religion, nationality, ethnicity, an

socioeconomic class. The revised instructions in this stud

asked respondents to consider himself/herself as a memb

of his/her important social groups such as a school class o

a work unit when answering questions because (1) the

groups are more salient in the daily life of general Chine

people and therefore more relevant to their evaluation oquality of life, and (2) previous research ndings (Kwa

Bond, & Singelis, 1997) suggested that the race-speci

form of CSES might not be good predictor of li

satisfaction. The scale developers (Luhtanen & Crocke

1992, Study 3) data, as well as those of others (Bettencou

& Dorr, 1997; Ethier & Deaux, 1990), showed th

modications to the reference groups used in the scale fo

research purposes did not negatively affect the reliability

the scale.

The CSES includes four 4-item subscales. The total sca

alpha is .88 (Luhtanen & Crocker, 1992). The Membersh

Esteem subscale (alpha 5 .75) assesses individual judgments of how worthy they are as members of the

social groups. A sample question is I am a worth

member of the group I belong to. The Private CS

subscale (alpha 5 .71) measures ones personal judgmen

of how good ones social group is. A sample question is

often regret that I belong to the group I do. The Publ

CSE subscale (alpha 5 .78) assesses ones judgments o

how positively other people evaluate ones social groups.

sample question is Overall, my group is considered goo

by others. The Importance to Identity subscale (alpha

.86) assesses the importance of ones social grou

memberships to ones self-concept. A sample question

Overall, my group memberships have very little to d

with how I feel about myself. In the present investigation

the coefcient alphas for Membership Esteem, Priva

CSE, Public CSE, and Identity Inuence subscales we

.51, .76, .67, and .48 respectively. A composite collectiv

self-esteem score was computed by averaging the items o

the four subscales and the coefcient alpha for this fu

scale was .83.

Procedure

The senior high school students and university studen

were invited to ll in a questionaire booklet (described a

86 ZHANG AND LEUNG

TABLE 1

Sample size, age range, and age mean in age and gender groups

Male Female Total

Sample Age Age Sample Age Age Sample Age Age

Age group size range mean size range mean size range mean

Middle school 205 1421 15.43 216 1523 15.39 423 1433 15.41

University 177 1824 20.77 179 1732 20.50 356 1732 20.63

Parent 155 4063 44.03 191 3574 42.52 348 3574 43.19

Grandparent 83 5488 70.47 103 4387 68.17 189 4388 69.11

Total 620 1488 31.47 689 1587 32.13 1316 1488 31.88


5/10

Immediately before starting the test the teachers were

trained in a 30-minute session on how to administrate the

questionnaire. They were told briey (1) the general

purpose of the test, (2) the requirement of the test, and (3)

the method of collecting questionnaires from parents and

grandparents. There were about 50 students in each class.

The time to nish the questionnaire booklet ranged from

28 to 52 minutes. After the test, the senior high school

students were given the same questionnaires for one of

their parents and one of their grandparents to complete at

a convenient time at home, and were asked to return the

completed booklets in 1 week. The teachers were given a

gift for their help in organizing the test. The students were

also given a gift for their contribution and the contribution

of their parents and grandparents.

To encourage truthful responses, participants were told

in the rst page of the booklet that their names were not

requested and that their answers would be kept

condential.

Because only one of the senior high school students

parents and only one of their grandparents were invited to

participate in the present investigation, the gender of

parent and grandparent participants was decided on a

random basis. In case the randomly assigned parent or

grandparent was not available for answering the

questionnaire booklet the students were asked to invite

another parent or grandparent to complete the booklet. If

both parents or both grandparents were unavailable, the

students were asked to return the unnished booklets to

their class teachers and to give a brief explanation. The

response rate was 84.31% for parents and 45.43% for

grandparents. According to the class teachers, the reasonsfor parents or grandparents not responding included (1)

they went out of the city for business; (2) they were too

busy; (3) the students did not live with their grandparents

and could not nd time to visit them within the week; (4)

they already passed away; and (5) they did not want to

respond.

The senior high school students, their parents, and their

grandparents represented three Chinese generations. The

university students are greatly exposed to modern ideas

and values in higher learning institutions. It is believed,

therefore, that these four age groups will provide useful

information on the overall picture of the relationshipbetween individual and collective self-esteem and life

satisfaction.

RESULTS

Correlation among variables

The means and standard deviations of life satisfaction and

self-esteem are presented in Table 2. The mean scores on a

7-pint Likert scale (06) were 3.34 for general life

satisfaction and 4.05 for life domain satisfaction, whichindicated a moderately positive evaluation. A correlation

analysis was also conducted for all variables and the

both self-esteem indicators were positively correlated wit

two satisfaction indicators. The correlation betwee

general life satisfaction and life domain satisfaction wa

signicant but moderate (r5 .46,p, .01). The correlatio

between individual self-esteem and collective self-esteem

was also moderate (r 5 .45, p , .01). These correlation

indicated that these two pairs of concepts were correlate

but different. The moderate correlation between the tw

self-esteems also suggested that there should be no seriou

multicollinearity problem if they were to be used as tw

independent variables in the regression analyses.

Effects of gender and age

Hierarchical regression analysis was used to te

moderating effects of gender and age on the relationshi

between the two self-esteems and two life satisfactions. Th

general procedure for testing moderating effects in th

present investigation was to enter predictors rst, then tenter moderators. The last step was to enter interactio

terms between predictors and moderators. A signican

interaction term would be taken as an indication of

moderating effect.

To test gender and age effects, individual self-esteem an

collective self-esteem were entered at Step 1. Gender an

age were entered at Step 2. In the last step, four interactio

terms, i.e., individual self-esteem 3 gender, collective sel

esteem 3 gender, individual self-esteem 3 age, an

collective self-esteem 3 age were together added into th

model. Table 3 and Table 4 present the results of th

regressions on general life satisfaction and life domaisatisfaction respectively.

At Step 1, individual self-esteem and collective sel

esteem were positively associated with two satisfactio

indices in two models.

At Step 2, age was positively related to two satisfactio

indices in two models, indicating that people in the old ag

group were more satised with their lives than people i

the young age group. Gender was not associated wit

general life satisfaction but was negatively associated wit

life domain satisfaction, indicating that the ma

participants were more satised with their life than the

female counterparts.At Step 3, there was a signicant interaction betwee

collective self esteem and gender in the general lif


TABLE 2

Means, standard deviations, and correlations of life satisfaction an

self-esteem indicators (N 5 1347)

1 2 3 4

1. General life satisfaction .46** .32** .268**

2. Life domain satisfaction .41** .36**

3. Individual self-esteem .45**

4. Collective self-esteem

Mean 3.34 4.05 3.96 3.96

SD 1.14 0.84 0.89 0.91

**p, .01.


6/10

indicated that the relationship between collective self-

esteem and general life satisfaction was stronger for the

male participants than for the female participants.

There were also signicant interactions between self-

esteems and demographic variables in the life domain

satisfaction model. The negative beta of individual self-

exteem 3 gender suggested that the inuence of individual

self-esteem on life domain satisfaction was stronger in the

male group than in the female group. The positive beta of

individual self-esteem3 age suggested that the inuence of

individual self-esteem on life domain satisfaction was

stronger in the older people than in the younger people.However, this pattern was reversed when collective self-

esteem was used as the predictor The negative beta of

collective self-esteem on life domain satisfaction wa

stronger in the younger people than in the older people.

DISCUSSION

Predictive power of gender and age

Gender effect on life satisfactionPrevious studies on gender difference of life satisfactio

have produced quite confusing results. On the one han

studies on positive wellbeing found that women reporte

greater happiness and life satisfaction than men (see th

review of Wood, Rhodes, & Whelan, 1989). A met

analysis (Haring, Stock, & Okun, 1984) indicated that me

were slightly happier than women, but the magnitude o

this difference was very small (mean r 5 .04). On the oth

hand, studies on negative affect and psychologic

symptomatology found that women reported having mo

emotional problems (see the review of Fujita, Diener,

Sandvik, 1991). Some researchers (Diener et al., 198

Warr, Barter, & Brownbridge, 1983; Wood et al., 198

proposed that positive affect and negative affect a

uncorrelated. One empirical study with multitrai

multimethod matrix analyses showed that positive affect

discriminable from negative affect (Lucas, Diener, & Su

1996). According to this understanding of affe

conceptualization, the gender difference of life satisfactio

that we found might be reasonable. However, oth

researchers (Diener et al., 1999; Fujita et al., 1991) tried t

use the gender difference on emotional intensity to expla

these confusing results. This notion argued that becauwomen had higher level of emotional intensity the

reported higher levels of both positive and negative affec

The result of this investigation indicated that the ma

participants had a higher level of life domain satisfactio

than their female counterparts. Since no negative affe

data was collected, it is impossible for this investigation t

test the emotional intensity hypothesis. However, there a

two other competing explanations. One is concerned wi

self-esteem, a source of life satisfaction. Kling, Hyd

Showers, and Buswell (1999) reported that in their met

analysis, which included 216 effect sizes and 97,12

respondents, males scored higher on standard measures global (individual) self-esteem than females, although th

difference was small. This lower level of self-esteem

women might lead to a lower level of life satisfactio

However, because both individual self-esteem effect an

collective self-esteem effect were controlled in the da

analysis by entering them into the two repression mode

in the rst step (Table 3), this explanation was not justie

Another competing explanation for womens lower lev

of life satisfaction is related to womens social status. I

Mainland China, although social status of women h

been greatly improved during the last half century,

general their social status is still lower than men. Generobservation tells us that there is serious discriminatio

toward female babies in rural areas Female graduat

88 ZHANG AND LEUNG

TABLE 3

Summary of hierarchical regression analysis for moderating effect of

age and gender on self-esteems general life satisfaction relation

(N = 1347)

Variable B SEB b Total R2

Step 1 .12

Individual self-esteem .32 .04 .25**

Collective self-esteem .19 .04 .15**

Step 2 .18

Gender 2.01 .06 2.01

Age .02 .00 .25**

Step 3 .19

Individual self-esteem 3 gender .05 .08 .11

Collective self-esteem 3 gender 2.20 .08 2.44**

Individual self-esteem 3 age 2.00 .00 2.25

Collective self-esteem 3 age .00 .00 .01

Gender: 1 5 men; 2 5 women

R2 5 .12 (p, .01) for Step 1;

DR2 5 .06 (p, .01) for Step 2;

DR2 5 .01 (p, .05) for Step 3.

**p, .01.

TABLE 4

Summary of hierarchical regression analysis for moderating effect of

age and gender on self-esteems life domain satisfaction relation

(N = 1347)

Variable B SEB b Total R2

Step 1 .21

Individual self-esteem .29 .03 .31**

Collective self-esteem .21 .03 .23**

Step 2 .22

Gender 2.10 .04 2.06*Age .00 .00 .10**

Step 3 .23

Individual self-esteem 3 gender 2.11 .06 2.31*

Collective self-esteem 3 gender 2.01 .06 2.02

Individual self-esteem 3 age .00 .00 .34*

Collective self-esteem 3 age 2.01 .00 2.61**

Gender: 1 5 men; 2 5 women

R2 5 .21 (p, .01) for Step 1;

DR2 5 .01 (p, .01) for Step 2;

DR2 5 .01 (p, .01) for Step 3.

*p, .05; **p , .01.


7/10

women who have been married take a greater share of

housekeeping work and child-care responsibility than their

husbands. Working women have less opportunity to be

promoted to a higher position than their male

counterparts. They make at least equal, and sometimes

more contribution to society, but they almost always live a

more difcult life than men. The unbalance between their

give to the society and their take from society might

make them more unsatised with their life, especially when

concrete criterions such as life domain satisfaction are

available for the judgment of life situations.

Age effect on life satisfaction

When no effect of age on life satisfaction indicators was

found in the study by Zhang and Leung (1999), the

authors speculated that the effect might be attenuated due

to the small age range of the participants (1229). This

speculation appeared to be true because age was associated

with life satisfaction indicators when the age range was

greatly enlarged (1488) in the present investigation. The

result indicated that older Chinese people had a higher

level of satisfaction than their offspring.

There are two different views on the relationship

between age and life satisfaction. Wilson (1967) proposes

that older people are less happy whereas Diener and

colleagues (1999) propose the opposite, i.e., older people

are happier. Wilsons view seems a reection of research

ndings before the 1970s whereas Diener et al.s view is

after the 1970s; apparently a more exact reection of

research development. Recent studies converge to show

that life satisfaction often increases, or at least does notdrop, with age (Herzog & Rodgers, 1981; Horley & Lavery,

1995; Larson, 1978; Stock, Okun, Haring, & Witter, 1983).

Cross-cultural studies based on representative samples

from multiple countries also show that life satisfaction

does not decline with age (Butt & Beiser, 1987; Ingelhart,

1990; Veenhoven, 1984). In a more recent study with

national probability samples of almost 60,000 adults from

40 nations, Diener and Suh (1998) found that there was a

slight upward trend in life satisfaction from people in their

20s to those in their 80s. The result of this investigation

was also consistent with this age trend after the 1970s.

Diener et al. (1999) suggested that the discrepancy ofcurrent ndings from those reviewed by Wilson may be

due to the fact that older people are now healthier and stay

involved in more life domains than did previous

generations (e.g., Bass, 1995).

This nding that life satisfaction is positively correlated

with age is also consistent with the multiple discrepancy

theory of satisfaction proposed by Michalos (1985). This

theory suggests that individuals compare themselves to

multiple standards including other people, past conditions,

aspirations and ideal levels of satisfaction, and needs or

goals. Satisfaction judgments are then based on

discrepancies between current conditions and thesestandards. The multiple discrepancy theory (Michalos,

1985) predicts that a discrepancy that involves an upward

will result in decreased satisfaction, wheres a downwar

comparison will result in increased satisfaction. As w

know, the older Chinese generation suffered from man

difculties and frustrations in Chinese modern history du

to World War II, the Chinese civil war, and a series o

political movements such as the Great Cultura

Revolution. It is not until the 1980s that Chinese peop

have enjoyed a relatively peaceful political environmen

and fast economic development. The sharp differences o

political, social, and economic environment before an

after the 1980s might help older Chinese generations hav

a greater ability to appreciate what they have than younge

Chinese generations. They might also be involved mor

often in downward comparison, i.e., in comparing th

dramatic political, social, and economic development wit

previous suffering situations stored in their bitter mem

ories. However, younger generations might be involved i

downward comparisons much less often because the

simply have no such bitter experience. They are gen

erations who have endured no hardship.

One support of this view comes from a study b

Cheung, Leung, Zhang, Song, and Xie (1996). In the

study on Chinese personality and social change, the

found in samples from two Chinese cities and one villag

that economic improvements associated with developmen

would lead to a better quality of living and more lif

satisfaction. Here city people might be involved more ofte

in downward comparisons than their village counterpart

because political, social, and economic development i

Chinese cities is more dramatic than in Chinese countr

areas.

Moderating effects of gender and age

Gender effect on self-esteemlife

satisfaction relationship

The results indicated that gender had moderating effec

on the relationship between collective self-esteem an

general life satisfaction and on the relationship betwee

individual self-esteem and life domain satisfaction. Thes

relationships were stronger for the male participant

General observation appears to show that men pay mor

attention to and spend more time on their careedevelopment whereas women pay more attention to an

spend more time on their family because of social ro

expectations and practice. It seems that collective sel

esteem is more important to men because work group o

study group are related to their career development, whic

is the first priority of their life according to soci

expectation. A higher level of collective self-esteem migh

imply a better social environment for their caree

development and therefore lead to a higher level of li

satisfaction. In comparison, women put more emphasis o

family life. When they dont have a good soci

environment to develop their career and dont have verhigh collective self-esteem, it might be easier for them t

turn their attention to their present or future family lif



8/10

accordance to womens social role according to Chinese

tradition (Hong, 1976). It also seems that individual self-

esteem is more important to men because it correlates

closely to personal achievements, which is considered a

major life goal for men according to Chinese tradition.

Recently, in a meta-analysis of 216 effect sizes in self-

esteem representing the testing of 97,121 respondents,

Kling et al. (1999) reported that the overall effect size was

.21, a small difference favouring males. Facing the

difculty of lower self-esteem, females might tend to

devalue the importance of self-esteem. Consequently the

relationship between individual self-esteem and life

satisfaction tended to be weaker in females.

Age effect on self-esteemlife

satisfaction relationship

The result also indicated that the relationship between

collective self-esteem and life domain satisfaction was

stronger for the younger participants, whereas the

relationship between individual self-esteem and life

domain satisfaction was stronger for the older

participants. The rst possible explanation of this result is

related to life task differences between younger and older

generations. The adolescents major life task is to learn.

They have to learn the means of survival from family and

society. Their development and wellbeing depend greatly

on the resources and help they can get from their social

environment. In comparison, the adults major life task is

to achieve social and civic responsibility. They have to

contribute to family and society. Their development and

wellbeing depend on how much they contribute. Thereforesocial support, including collective support, seems more

critical for survival and wellbeing for adolescents, which

leads to a stronger relationship between collective self-

esteem and life satisfaction. Individual contribution,

individual responsibility, and individual achievement

seems more critical for survival and wellbeing for older

people, which leads to a stronger relationship between

individual self-esteem and life satisfaction. This result

seems consistent with previous ndings, which indicated

that individual self-esteem was the best predictor of life

satisfaction in the elderly (Campbell et al., 1976), but

parent relationship was more powerful than individualself-esteem in predicting life satisfaction in adolescents

(Huebner, Gilman, & Laughlin, 1999; Leung & Leung,

1992). According to this reasoning, the moderating effect

of age on the relationship between self-esteems and

general life satisfaction has also been expected. However,

there is no available explanation as to why this

investigation only found the moderating effect of age for

the relationship between self-esteems and life domain

satisfaction but not for the relationships between self-

esteems and general life satisfaction.

The second possible explanation is related to the

working status of our sample. Most of the grandparents inour sample have already retired (in Mainland China,

retirement policy is age 60 for men and 55 for women)

rest of their life. The importance of work groups decreas

as soon as they retire. Therefore, when work or stud

groups were taken as reference groups in our collectiv

self-esteem test, the relationship between collective sel

esteem and life satisfaction was demonstrated to be weak

in the older people than in the younger people.

The present investigation indicated that gender and ag

did moderate the relationship between two self-estee

indicators and two life satisfaction variables. The

moderating effects seem to reect the life task differenc

in different genders at different ages among Mainlan

Chinese. Additional research is needed with other types

samples, especially with samples from some individuali

cultures, to see whether the results could generalize t

these cultures.

Manuscript received February 20

Revised manuscript received August 20

REFERENCES

Andrews, F.M., & Crandall, R. (1976). The validity of measurof self-reported wellbeing. Social Indicators Research, 3, 11

Andrews, F.M., & Withey, S.B. (1976). Social indicators of webeing: Americans perceptions of life quality. New YorPlenum Press.

Bass, S.A. (1995). Older and active: How Americans over 55 acontributing to society. New Haven, CT: Yale University Pre

Baumeister, R.F. (1998). The self. In D.T. Gilbert, S.T. Fiske,G. Lindzey (Eds.), The handbook of social psychology (4th edpp. 680740). Boston, MA: The McGraw-Hill Companies.

Bettencourt, B.A., & Dorr, N. (1997). Collective self-esteem as

mediator of the relationship between allocentrism and subjetive wellbeing. Personality and Social Psychology Bulletin, 2955964.

Butt, D.S., & Beiser, M. (1987). Successful aging: A theme finternational psychology. Psychology and Aging, 2, 8794.

Campbell, A. (1981). The sense of wellbeing in America: Recepatterns and trends. New York: McGraw-Hill.

Campbell, A., Converse, P.E., & Rodgers, W.L. (1976). The quaity of American life. New York: Russell Sage Foundation.

Cheung, F.M., Leung, K., Zhang, J.X., Song, W.Z., & Xie, D(1996). Chinese personality and social change. In MBrosseau, S. Pepper, & S.K. Tsang (Eds.), China review (p343366). Hong Kong: Chinese University Press.

Crocker, J., Luhtanen, R., Blaine, B., & Broadnax, S. (1994Collective self-esteem and psychological wellbeing amonwhite, black, and Asian college students. Personality anSocial Psychology Bulletin, 20, 503513.

Crocker, J., & Major, B. (1989). Social stigma and self-esteemThe self-protective properties of stigma. Psychological Revie96, 608630.

Diener, E. (1984). Subjective wellbeing. Psychological Bulletin, 9542575.

Diener, E. (2000). Subjective wellbeing: The science of happineand a proposal for a national index. American Psychologis55, 3443.

Diener, E., & Diener, M. (1995). Cross-cultural correlates of lisatisfaction and self-esteem. Journal of Personality and SociPsychology, 68, 653663.

Diener, E., Emmons, R.A., Larsen, R.J., & Grifn, S. (1985). T

Satisfaction With Life Scale. Journal of PersonaliAssessment, 49, 7175.

Diener, E., & Suh, E.M. (1998). Age and subjective wellbeing: A

90 ZHANG AND LEUNG


9/10

Diener, E., Suh, E.M., Lucas, R.E., & Smith, H.L. (1999) .Subjective wellbeing: Three decades of progress. PsychologicalBulletin, 125, 276302.

Ethier, K., & Deaux, K. (1990). Hispanics in ivy: Assessing iden-tity and perceived threat. Sex Roles, 22, 427440.

Feist, G.J., Bodner, T.E., Jacobs, J.F., Miles, M., & Tan, V. (1995).Integrating top-down and bottom-up structural modes ofsubjective wellbeing: A longitudinal investigation. Journal ofPersonality and Social Psychology, 68, 138150.

Fujita, F., Diener, E., & Sandvik, E. (1991). Gender differencesin negative affect and wellbeing: The case for emotionalintensity. Journal of Personality and Social Psychology, 61,427434.

Guo, J.W. (1992). A multidimensional study of the life satisfac-tion index in urban elderly. Acta Psychologica Sinica, 24,2834.

Haring, M.J., Stock, W.A., & Okun, M.A. (1984). A research syn-thesis of gender and social class as correlates of subjectivewellbeing. Human Relations, 37, 645657.

Herzog, A.R., & Rodgers, W.L. (1981). Age and satisfaction:Data from several large surveys. Research on Aging, 3,142165.

Hong, L.K. (1976). The role of women in the Peoples Republicof China: Legacy and change. Social Problems, 23, 545557.

Horley, J., & Lavery, J.J. (1995). Subjective wellbeing and age.Social Indicators Research, 34, 275282.

Huebner, E.S., Gilman, R., & Laughlin, J.E. (1999). A multi-method investigation of the multidimensionality of childrenswellbeing reports: Discriminant validity of life satisfactionand self-esteem. Social Indicators Research, 46, 122.

Inglehart, R. (1990). Culture shift in advanced industrial society.Princeton, NJ: Princeton University Press.

Ji, Y.F., & Yu, X. (1993). The assessment of self-esteem. In X.D.Wang (Ed.), Rating scales for mental health (pp. 251252).Beijing: Journal of Chinese Mental Health.

Kling, K.C., Hyde, J.S., Showers, J.C., & Buswell, B. (1999).Gender differences in self-esteem: A meta-analysis.Psychological Bulletin, 125, 470500.

Kwan, V.S.Y., Bond, M.H., & Singelis, T.M. (1997). Panculturalexplanations for life satisfaction: Adding relationship har-mony to self-esteem. Journal of Personality and SocialPsychology, 73, 10381051.

Lance, C.E., Lautenschlager, G.J., Sloan, C.E., & Varca, P.E.(1989). A comparison between bottom-up, top-down, andbidirectional models of relationships between global and lifefacet satisfaction. Journal of Personality, 57, 601624.

Larsen, R.J., Diener, E., & Emmons, R.A. (1985). An evaluationof subjective wellbeing measures. Social Indicators Research,17, 117.

Larson, R. (1978). Thirty years of research on the subjective well-being of older Americans. Journal of Gerontology, 33, 109125.

Leung, J.P., & Leung, K. (1992). Life satisfaction, self-concept,

and relationship with parents in adolescence. Journal of Youthand Adolescence, 21, 653665.

Lucas, R.E., Diener, E., & Suh, E. (1996). Discriminant validity

of wellbeing measures. Journal of Personality and SociPsychology, 71, 616628.

Luhtanen, R., & Crocker, J. (1992). A collective self-esteem scalSelf-evaluation of ones social identity. Personality and SociPsychology Bulletin, 18, 302318.

Michalos, A.C. (1985). Multiple discrepancies theory (MDTSocial Indicators Research, 16, 347413.

Nehrke, M.F., Hulicka, I.M., & Morganti, J.B. (1980). Age diferences in life satisfaction, locus of control, and self-concep

International Journal of Aging and Human Development, 12533.Neto, F. (1993). The satisfaction with life scale: Psychometri

properties in an adolescent sample. Journal of Youth anAdolescence, 22, 125134.

Rosenberg, M. (1965). Society and the adolescent self-imagPrinceton, NJ: Princeton University Press.

Scarpello, V., & Campbell, J.P. (1983). Job satisfaction: Are all thparts there? Personnel Psychology, 36, 577600.

Seligman, M.E.P., & Csikszentmihalyi, M. (2000). Positive pschology: An introduction. American Psychologist, 55, 514.

Shin, D.C., & Johnson, D.M. (1978). Avowed happiness as aoverall assessment of the quality of life. Social IndicatoResearch, 5, 475492.

Stock, W.A., Okun, M.A., Haring, M.J., & Witter, R.A. (1983Age and subjective wellbeing: A meta-analysis. In R.J. Ligh(Ed.), Evaluation studies: Review annual, Vol. 8 (pp. 279302Beverly Hills, CA: Sage.

Tan, H.P. (1998). Compilation of a mental health scale for midle school students. Psychological Science China, 21, 42943

Veenhoven, R. (1984). Conditions of happiness. Dordrecht, ThNetherlands: D. Reidel Publishing.

Veenhoven, R. (1996). Developments in satisfaction-researcSocial Indicators Research, 37, 146.

Warr, P., Barter, J., & Brownbridge, G. (1983). On the indepenence of positive and negative affect. Journal of Personality anSocial Psychology, 44, 644651.

Wilson, W. (1967). Correlates of avowed happiness. PsychologicBulletin, 67, 294306.

Wood, W., Rhodes, N., & Whelan, M. (1989). Sex differences positive well-being: A consideration of emotional style anmarital status. Psychological Bulletin, 106, 249264.

Xiang, M.J., Wu, X.G., & Liu, X.H. (1995). Life satisfaction anfactors related to elderly citizens in Beijing. Acta PsychologicSinica, 27, 395399.

Xu, H.L. (1994). A study on life satisfaction of the elderly and ipredictors. Chinese Journal of Mental Hygiene, 8, 160162.

Yang, Y.C. (1988). Investigation on life satisfaction of the elderand social-psychological factors. Chinese Journal of MentHygiene, 2, 912.

Zhang, L.W., & Leung, J.P. (1999). Added value of collective selesteem beyond individual self-esteem and social support in prdicting athletes general and career satisfaction. Pap

presented for the Third International Congress of AsiaSouth Pacic Association of Sport Psychology. Wuhan, PChina, October.



10/10

Moderating Effects of Gender and Age on the Relationship Between Self-esteem and Life Satisfaction...

Documents

Transcript of Moderating Effects of Gender and Age on the Relationship Between Self-esteem and Life Satisfaction...