Professional Documents
Culture Documents
Shek - MIDLIFE CRISIS IN CHINESE MEN AND WOMEn - 1996
Shek - MIDLIFE CRISIS IN CHINESE MEN AND WOMEn - 1996
net/publication/14582770
CITATIONS READS
23 3,315
1 author:
Daniel Shek
The Hong Kong Polytechnic University
966 PUBLICATIONS 13,024 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Daniel Shek on 07 April 2015.
130(1), 109-119.
Chinese Midlife Crisis Scale. An examination of the levels of concerns in six midlife age
MIDLIFE CRISIS IN CHINESE MEN AND WOMEN groups revealed that, although some respondents were dissatisfied with their own work and
personal achievement, a majority did not respond in the crisis direction, hence the findings
Daniel T.L. Shek do not lend strong support to the existence of a normative midlife crisis. The findings also
Department of Social Work revealed that, although midlife crisis levels were different in the various age groups, there
The Chinese University of Hong Kong was no clear rise or peak in concerns in any particular age group. Further analyses based on
different dimensions of midlife crisis showed that women displayed higher levels of overall
This work was financially supported by the Research Grants Council of the University and
Shek, Department of Social Work, The Chinese University of Hong Kong, Shatin, Hong
Kong.
MIDLIFE CRISIS IN CHINESE MEN AND WOMEN In their discussion of the nature of midlife transition, Cytrybaum et al. (1980)
suggested that there is a need to differentiate between developmental processes and outcomes
A review of the Western literature shows that relatively few attempts have been in midlife development. While triggers (such as shift in time orientation and illness of
conducted to study the nature of middle adulthood (midlife). As commented by Hunter and parents) and developmental tasks (such as reorientation to work and reassessment of primary
Sundel (1989), systematic research began only about 15 years ago. Nevertheless, in spite of relationships) are central to the developmental processes, midlife-related symptoms (such as
the lack of scientific data, cultural myths and stereotypes are numerous. In particular, the alcoholism and mental disorders) should be seen as outcomes of midlife transition.
portrayal of people in middle adulthood have a "midlife crisis" often appears in the press and With specific reference to the developmental processes involved in midlife transition
media (Chew, 1976; Conway, 1980; Nichols, 1986). (Farrell & Rosenberg, 1981; Hunter & Sundel, 1989; McCrae & Costa, 1984), several
The term "midlife crisis" was originally coined by Jaques (1965) who, on the basis of components have been proposed to be intrinsic to the conceptualization and assessment of
his study of the personal experiences of artists, concluded that midlife people encountered a midlife crisis: a) worries about the future; b) inability to enjoy leisure time; c) a feeling that
period of crisis which was triggered by the realization of their own mortality and change in health is deteriorating; d) a negative evaluation of the parent-child relation; e) a negative
time frame from "time since birth" to "time left to live". The idea of midlife crisis or midlife evaluation of the marital relation; f) a negative evaluation of the work life; g) existential
transition was subsequently proposed in several models on adult development (Gould, 1978; confusion, such as confused life meaning, lack of time to achieve and lack of achievement;
Levinson, 1978; Sheehy, 1976, 1981). In these models, it is suggested that a midlife person and h) stress arising from taking care aged relatives.
would be increasingly aware of mortality and the time left, and in response to such an Probably because of the existence of different definitions of the term, empirical
awareness, he or she would evaluate his or her life, particularly with respect to one's family, support for the notion of midlife crisis is equivocal. On one hand, there is evidence that
work and existence. people experience midlife crisis: Levinson (1978) found that 80% of middle-aged subjects
Although the concept of midlife crisis appears to be elegant, it is beset with had such a crisis; Ciernia (1985) reported that 70% of male midlife people expressed that
conceptual difficulties and there are wide variations in the conception of the nature of midlife they had a crisis; Tamir (1989) suggested that midlife transition was characterized by
crisis (e.g., Cytrybaum et al., 1980; Sutherland, 1989; Whitbourne, 1986). Operationally, "deep-seated self-doubts or confusion" (p.161) which was quite prevailing.
different indicators have been used to measure the concept of midlife crisis. In the Midlife In contrast, there are data not supporting the notion of midlife crisis. Although it was
Crisis Scale used by McCrae and Costa (1984), items covering "sense of meaninglessness, proposed that people would re-examine their lives around midlife, Vaillant (1977) suggested
dissatisfaction with job and family, inner turmoil and confusion, and sense of impending that traumatic crisis amongst non-clinical midlife people was rare. McCrae and Costa (1984)
physical decline and death" (p.103-104) were employed. In the Midlife Crisis Scale devised used the Midlife Crisis Scale to study midlife crisis in male men and concluded that there
by Farrell and Rosenberg (1981), similar items on the evaluation of marriage, children and was no evidence of a peak of distress or midlife characteristics. Based on the responses of
work as well as existential confusion were included. 500 men to a similar Midlife Crisis Scale, Farrell and Rosenberg (1981) found no significant
differences in midlife crisis levels amongst subjects in different age groups. Epidemiological whether midlife crisis also exists in Chinese people and whether the related phenomena are
studies also do not show a piling up of psychiatric disorders and maladjustment in midlife similar to the Western culture. It can be conjectured that Chinese people's emphasis on
years (Ciernia, 1985; Hunter & Sundel, 1989; Pearlin, 1975). In a review of the evidence on familism, interpersonal harmony, and fatalism might make the occurrence of midlife crisis
midlife crisis, Chiriboga (1989) concluded that "only 2 percent to 5 percent of middle agers" less prominent in Chinese societies. Furthermore, empirical investigation of the nature of
(p.117) had serious midlife problems. midlife adjustment in the Chinese context is important because the Chinese in mainland
There are wide variations in the methods used to assess midlife crisis. While China and Hong Kong constitute roughly one-fifth of the world's population, which suggests
interviews were mainly used by Levinson (1978) and Vaillant (1977), Gould (1978) and that the generalizability of the related theoretical accounts and hypotheses can be tested.
Sheehy (1981) used questionnaires. Amongst those studies in which questionnaires or In contrast to the lack of empirical evidence on the notion of "midlife crisis", there
psychological scales were used, there are three associated problems. First, information on the are prevailing cultural stereotypes concerning midlife in the Chinese culture. Some negative
psychometric properties of the instruments have seldom been reported (e.g., Farrell & Chinese cultural beliefs about midlife include: "ren dao zhong nien wan shih ai" (when a
Rosenberg, 1981; McCrae & Costa, 1984). Second, an examination of the tools used in some person reaches the midlife years, everything is sad); "ren dao zhong nien wan shih you"
of the studies suggests that such tools might be problematic. An analysis of the items in the (when a person reaches the midlife years, everything is worrying). In view of such cultural
Midlife Crisis Scale (Farrell & Rosenberg, 1981) showed that there is a strange mixture of beliefs, it would be interesting to examine the validity of such cultural stereotypes. In a
general attitudinal statements (e.g., item 2: "many men I know are undergoing what you pioneer study of midlife crisis in Chinese people, Shek, Lam, Tsoi, and Lam (1993)
would call a change in life or a middle-age identity crisis") and statements which have developed the Chinese Midlife Crisis Scale (C-MCS). Shek (1994) reported data supporting
personal relevance (e.g., item 4: "I am still finding new challenges and interest in my work"). the reliability and concurrent validity of the C-MCS and proposed that three dimensions
In fact, a visual inspection of the items suggests that the items included may lack construct (Self-Related Midlife Problems, Other-Related Midlife Problems and Fear of Aging) are
validity. Finally, among those studies in which a midlife crisis scale was used, the concept of intrinsic to the scale.
crisis has been generally regarded as a unidimensional concept. However, it is legitimate to The present study has several purposes. First, it reports empirical evidence on the
ask whether there are different dimensions underlying the concept of midlife crisis (i.e., responses of Chinese midlife people to the Chinese Midlife Crisis Scale and the related
multi-dimensional concept) or it is a unidimensional construct. responses would then be used as the basis to evaluate the existence of a "normative" midlife
A review of the literature also shows that existing studies have been conducted in crisis in Chinese people. If midlife crisis is a normative developmental experience, it would
Western societies and no effort has to date been made to examine this problem in the Chinese be expected that a majority of the respondents in one or more of the midlife age groups
culture. Among the 3,548 bibliographic citations pertinent to Chinese patterns of behavior by would respond to the C-MCS Scale in the crisis direction. Secondly, with reference to some
Ho, Spinks, and Yeung (1989), there is only one citation related to the subject "middle aged" of the proposals that there is a definite period of midlife adjustment or crisis (e.g., Levinson,
and no reported study on "midlife crisis". From a cross-cultural perspective, it is useful to ask 1978), age differences in midlife crisis have been examined. Finally, since few studies have
been conducted to examine gender differences in midlife crisis (Hunter & Sundel, 1989; METHOD
Reinke, Holmes, & Harris, 1985), the paper examines gender differences in midlife crisis in
This study is part of a general study of the adjustment and mental health of Chinese The Chinese Midlife Crisis Scale (Shek, 1994) was used to assess midlife crisis (see
married adults. Because data on the mental health and other areas of adjustment of the Appendix). The distribution of the C-MCS items with reference to different areas of midlife
participants have been discussed elsewhere (Shek, 1994; Shek, in press), the present paper crisis is as follows: worries about aging (item 4 and item 8); inability to enjoy leisure time
focuses primarily on midlife crisis in Chinese married adults. (item 1); a feeling that health is deteriorating (item 15); a negative evaluation of the
parent-child relation (item 2, 6 and 14); a negative evaluation of the marital relation (item 13);
a negative evaluation of the work life (items 7, 9 and 11); existential confusion, such as
confused life meaning (item 12), lack of time to achieve (item 10) and lack of achievement
(item 5); and stress arising from taking care the aged relatives (item 3). Shek (1994) found
that three factors, Self-Related Midlife Problems (items 1, 5, 7, 9, 10, 11, and 15),
Other-Related Midlife Problems (items 2, 3, 6, 12, 13 and 14), and Fear of Aging (items 4
As indicated, the data reported were from a study that was designed to investigate the
psychological well-being and adjustment of married adults in Hong Kong. Therefore, the
participants were also requested to respond to other scales that assess different aspects of
mental health and adjustment. These include the Chinese versions of the General Health
Children Scale and Parent-Child Relational Quality Scale. Shek (in press) showed that the
The participants were 1,501 married adults aged between 30 to 60 years. They were
selected from the public housing estates in Hong Kong by the multiple-stage cluster
sampling method (Moser & Kalton, 1980). Since the sample was drawn from public housing
estates, their socioeconomic backgrounds were largely middle and lower-middle class. Each RESULTS
containing the assessment tools was used to elicit the responses from the subject. The whole Consistent with some of the previous practices (Reinke, Holmes, & Harris, 1985), the
interview process took roughly one hour to complete. respondents were classified into six groups as follows: Group 1 (age 30 to age 34.9, n=157),
With reference to some of the existing frameworks on human development (e.g., Group 2 (age 35 to age 39.9, n=300), Group 3 (age 40 to age 44.9, n=345), Group 4 (age 45
Havighurst, 1972; Hurlock, 1980), the participants could be treated as persons in early to age 49.9, n=296), Group 5 (age 50 to age 54.9, n=210) and Group 6 (age 55 to age 60,
adulthood (commonly regarded as the period between 20 to 40 years of age) and middle n=193).
adulthood (the period between 40 to 60). However, if we adopt the idea of other researchers Liberal as well as conservative criteria were used to examine the levels of the crisis in
who have asserted that middle adulthood begins at the twenties (Erikson, 1963) or thirties these groups. According to the liberal criterion, responses indicating agreement to the
(Gould, 1978; Sheehy, 1976), the participants may be regarded as midlife participants in the negatively-worded items ("Slightly Agree", "Moderately Agree" and "Strongly Agree") and
broadest sense of the term. responses indicating disagreement to the positively-worded items ("Slightly Disagree",
"Moderately Disagree" and "Strongly Disagree") were regarded as responses in the crisis
direction. Concerning the conservative criterion, only the "Moderately Agree" and "Strongly
Agree" responses to the negatively-worded items and the "Moderately Disagree" and
the crisis direction. Using both criteria, the total percentage of crisis responses for each
The above analysis generally show two phenomena. First, while slightly more than
half of the respondents in the various age groups indicated that they had problems in the
areas of work (items 9 and 11), personal achievement (items 5 and 10) and health (item 15) if
the liberal criterion is used, there is no item (except item 15 in Group 6) to which more than
half of the respondents in the various age groups responded in the crisis direction if the
conservative criterion is adopted. Secondly, the data did not appear to show that midlife
crisis concerns clustered at any particular age. One-way ANOVAs (with age as the
independent variable and the total C-MCS and its subscale scores as dependent variables)
were further performed to examine whether there were differences in crisis levels amongst
the age groups. The results showed that while there were significant differences amongst the carried out. The MANOVA results were significant for the above-mentioned groups of
six age groups on the Self-Related Midlife Problems subscale and total midlife crisis scores, combined dependent variables (Group 1: F(15/698)=5.03, p < .0001 for random subsample 1
the data do not show any clear rise or peak in midlife crisis level in any particular age group. and F(15/694)=14.3, p < .0001 for random subsample 2; Group 2: F(3/710)=11.24, p < .0001
Further details of the percentage data based on the conservative and liberal criteria and the for random subsample 1 and F(3/706)=6.20, p < .0001 for random subsample 2). These
ANOVA data are available upon request. analyses consistently show that women displayed a higher level of midlife concerns as
A series of t-tests was initially performed to evaluate gender differences on the compared to men.
various measures of midlife crisis. The results generally showed that females displayed more
midlife crisis problems. The results are summarized in Table 1. The effect sizes involved
were also calculated according to the procedures described in Smith and Glass (1977).
-------------------------------
--------------------------------
Since the employment of a series of t-test might increase the chance of committing
Type 1 error, further analyses using multivariate analysis of variance (Norusis, 1990) was
carried out for the following groups of dependent variables: a) 15 item scores of the C-MCS;
and b) three subscale scores of the C-MCS. With the use of Wilk's criterion, significant
multivariate effects of gender were found for all groups of combined dependent variables,
generally reveal a significant association between gender and the various categories of
midlife crisis measures and they showed that women appeared to display a higher level of
midlife concerns. Since the omnibus multivariate F's were significant, univariate F statistics
were further included as aids to interpret the significant MANOVA results and to clarify the
relationship between the independent variable and individual dependent variables. The
univariate ANOVA results were consistent with the original t-test analyses.
The stability of the MANOVA results was further tested by splitting the total sample
into two random subsamples (subsample 1 and subsample 2) and identical procedures were
DISCUSSION that age 40 to age 45 is the period during which intensive midlife crisis occurs. Of course, in
view of the cross-sectional nature of the research design, the associated data should be
If midlife crisis is a normative developmental phenomenon which occurs at some interpreted with caution. In addition, since the present analysis centres around chronological
time within the midlife period, it is reasonable to expect that a majority of midlife age without reference to the different stages of the family life cycle, its limitation should be
participants in at least one of the age groups should respond to the C-MCS items in the realized. For the problem of gender differences, two aspects of the data require attention. The
"crisis" direction. However, the present data do not appear to support this expectation. If the first aspect is that while married females in general displayed more midlife crisis-related
conservative criterion was used, there was no item to which more than half of the participants concerns than males, females were found to be more satisfied with their work (item 7). This
in any age group responded in the crisis direction (except item 15 in Group 6). Even if the phenomenon might be due to the possibility that, since Chinese married women tend to place
liberal criterion was adopted, only slightly more than half of the participants displayed more emphasis on their marriages than work, they may set a lower expectation in their work.
work-related problems (items 9 and 11) and frustration about personal achievement (items 5 As a result, it may be easier for married women to derive more satisfaction from their work
and 10). The analyses of variance data further show that although there are significant than married men. The second aspect is that, while gender differences in midlife crisis
differences amongst different age groups in the midlife period, there is no sign indicating that concerns are not related to Self-Related Midlife Problems, significant gender differences on
the level of midlife crisis is highest in a particular age group. Taken as a whole, the data do Other-Related Midlife Problems, Fear of Aging and overall midlife crisis were observed.
not seem to support the proposal that midlife crisis is a normative developmental This finding suggests that different dimensions of midlife crisis must be taken into account
phenomenon. They also do not give strong support to the popular Chinese sayings of "ren when gender differences in midlife crisis are assessed.
dao zhong nien wan shih ai" (when a person reaches the midlife years, everything is sad) and The data on the significant gender differences in midlife concerns are consistent with
"ren dao zhong nien wan shih you" (when a person reaches the midlife years, everything is the general finding that married females tend to have a poorer psychological well-being, such
worrying). as depression and psychiatric morbidity, than males (e.g., Bernard, 1972; Gove, Hughes, &
The above findings are consistent with some of the previous data in the literature. In Style, 1983). The observation that women had more problems with their children and were
an attempt to compare different midlife groups with pre- and post-midlife groups according more inclined to choose to be single suggests that midlife married women, as compared to
to different definitions of midlife, McCrae and Costa (1984) found "no evidence at all of a midlife married men, may not be so satisfied with their family life. This finding is consistent
peaking of midlife concerns at any age in our range" and concluded that "at any given time, with Bernard's (1972) claim that married women tend to derive less advantages from their
only a small percentage of men are in a crisis, and they are not likely to cluster at any marital lives.
particular age" (p.104). Similarly, Farrell and Rosenberg (1981) showed no significant In evaluating gender differences involved in midlife concerns, it should be realized
differences amongst men in different midlife periods in terms of their midlife crisis scores. that the associated effect sizes were in the "low" range (Cohen, 1988). Nevertheless, in view
On the other hand, the present findings do not seem to support Levinson's (1978) proposal of the paucity of research data on midlife development in females, the data in the present
2. Chew, P. (1976). The inner world of the middle-aged man. New York: McMillan.
3. Chiriboga, D.A. (1989). Mental health at the midpoint: Crisis, challenge, or relief? In S.
Hunter & M. Sundel (Eds.), Midlife myths, issues, findings, and practice implications
(pp.116-194). Newbury Park, CA: Sage.
4. Ciernia, J.R. (1985). Myths about male midlife crises. Psychological Reports, 56,
1003-1007.
5. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.).
Hillsdale, NJ: Erlbaum.
6. Conway, S. (1980). You and your husband's midlife crisis. New York: New American
Library.
7. Cytrybaum, S., Blum, L., Patrick, R., Stein, J., Wadner, D., & Wilk, C. (1980). Midlife
development: A personality and social systems perspective. In L. Poon (Ed.), Aging in
the 1980s (pp.463-474). Washington, DC: American Psychological Association.
9. Farrell, M.P., & Rosenberg, S. (1981). Men at midlife. Boston: Auburn House.
11. Gove, W.R., Hughes, M., & Style, C.B. (1983). Does marriage have positive effects on
the individual? Journal of Health and Social Behavior, 24, 122-131.
12. Havighurst, R.J. (1972). Developmental tasks and education. New York: McKay.
13. Ho, D.Y.F., Spinks, J.A., & Yeung, C.S.H. (Eds.) (1989). Chinese patterns of behavior: 24. Sheehy, G. (1976). Passages: Predictable crises of adult life. New York: E.P. Dutton.
A sourcebook of psychological and psychiatric studies. New York: Praeger.
25. Sheehy, G. (1981). Pathfinders. New York: Morrow.
14. Hunter, S., & Sundel, M. (1989). Midlife myths: Issues, findings and practice
implications. Newbury Park, CA: Sage Publications. 26. Shek, D.T.L. (1994). A scale for the assessment of midlife crisis in Chinese people.
Manuscript submitted for publication.
15. Hurlock, E.B. (1980). Developmental psychology: A life-span approach. New Delhi:
Tata McGraw-Hill Publishing Company. 27. Shek, D.T.L. (in press). Marital quality and psychological well-being of married adults
in a Chinese context. Journal of Genetic Psychology.
16. Jaques, E. (1965). Death and the mid-life crisis. International Journal of Psychoanalysis,
46, 502-514. 28. Shek, D.T.L., Lam, M.C., Tsoi, K.W., & Lam, C.M. (1993). Marital adjustment,
parent-child relations, and psychological well-being of midlife married adults in Hong
17. Levinson, D.J. (1978). The seasons of a man's life. New York: Knopf. Kong. Hong Kong Journal of Social Work, 27(2), 57-58.
18. McCrae, R.R., & Costa, P.T.J. (1984). Emerging lives, enduring dispositions: 29. Smith, M.L., & Glass, G.V. (1977). Meta-analysis of psychotherapy outcome studies.
Personality in adulthood. Boston: Little, Brown. American Psychologist, 32, 752-760.
19. Moser, C.A., & Kalton, G. (1980). Survey methods in social investigation. London: 30. Sutherland, S. (1989). McMillan dictionary of psychology. London: McMillan Press.
Heinemann Educational Books.
31. Tamir, L.M. (1989). Modern myths about men at midlife: An assessment. In S. Hunter
20. Nichols, M.P. (1986). Turning forty in the eighties: Personal crisis, time for change. & M. Sundel (Eds.), Midlife myths: Issues, findings and practice implications
New York: Norton. (pp.157-180). Newbury Park, CA: Sage.
21. Norusis, M.J. (1990). SPSS advanced statistics user's guide. Chicago: SPSS 32. Vaillant, G.E. (1977). Adaptation to life. Boston: Little, Brown.
Corporation.
33. Whitbourne, S.K. (1986). Adult development. New York: Praeger.
22. Pearlin, L. (1975). Sex roles and depression. In N. Datan & L. H. Ginsberg (Eds.), Life
span developmental psychology: Normative life crises (pp.191-207). New York:
Academic Press.
23. Reinke, B.J., Holmes, D.S., & Harris, R.L. (1985). The timing of psychosocial changes
in women's lives: The years 25 to 45. Journal of Personality and Social Psychology,
48(5), 1353-1364.
APPENDIX TABLE 1
View publication stats
Items of the Chinese Midlife Crisis Scale Mean differences between male and female participants on the various measures of midlife
crisis
MALE FEMALE
Item 1: I know how to use leisure time
Variable Mean S.D. Mean S.D. t-value Effect Size
Item 2: I am increasingly aware that it is difficult for me to interact with my children
Item 1 2.24 1.17 2.16 1.23 1.32 ns -
Item 3: I encounter great stress in taking care of my aged relatives Item 2 2.61 1.55 2.98 1.66 -4.46 *** .22
Item 3 2.97 1.80 2.96 1.89 0.09 ns -
Item 4: I do not worry that I will have financial difficulty when I get old Item 4 3.17 1.74 3.40 1.79 -2.53 * .13
Item 5 3.46 1.78 3.35 1.77 1.26 ns -
Item 5: I am disappointed that given my age, I still have not had much achievement Item 6 1.90 0.96 2.04 1.01 -2.85 ** .14
Item 7 2.91 1.48 2.60 1.43 4.18 *** .22
Item 6: I am satisfied with my children Item 8 2.97 1.71 3.20 1.76 -2.51 * .13
Item 9 3.58 1.78 3.73 1.72 -1.63 ns -
Item 7: I am satisfied with my present work Item 10 3.45 1.72 3.48 1.75 -0.41 ns -
Item 11 3.76 1.96 3.71 1.89 0.49 ns -
Item 8: I do not worry about my life in old age Item 12 2.50 1.61 2.70 1.73 -2.33 * .12
Item 13 1.81 1.42 2.30 1.78 -5.87 *** .28
Item 9: I am increasingly aware that my work is boring and monotonous Item 14 1.86 1.41 2.28 1.71 -5.23 *** .25
Item 15 3.91 1.63 3.98 1.63 -0.80 ns -
Item 10: I feel that I do not have enough time to complete those things which I desire to complete
Subscale 1 23.33 6.45 23.00 6.35 0.99 ns -
Item 11: If I can live again, I will choose my present job Subscale 2 13.62 4.89 15.24 5.88 -5.71 *** .28
Subscale 3 6.14 3.10 6.60 3.23 -2.81 ** .14
Item 12: I am increasingly uncertain why I exist Total 43.16 10.88 44.85 11.78 -2.80 ** .14
Item 14: If I can live again, I will still choose to have children Subscale 1: Self-Related Midlife Problems scores. Subscale 2: Other-Related Midlife
Problems scores. Subscale 3: Fear of Aging scores. Total: total C-MCS scores (Shek, 1994).
Item 15: My health is gradually deteriorating
Note: Items 2, 3, 5, 9, 10, 12, 13 and 15 are negatively-worded *** p < .001
Items 1, 4, 6, 7, 8, 11, and 14 are positively-worded ** p < .01
* p < .05
ns non-significant