Content uploaded by Iva Poláčková Šolcová
Author content
All content in this area was uploaded by Iva Poláčková Šolcová on Feb 16, 2016
Content may be subject to copyright.
For the Benefit of Others: Generativity and Meaning in Life in the Elderly
in Four Cultures
Jan Hofer and Holger Busch
University of Trier Alma Au
Hong Kong Polytechnic University
Iva Polácˇková Šolcová
The Academy of Sciences of the Czech Republic Peter Tavel
Palacky University
Teresa Tsien Wong
Hong Kong Polytechnic University
Generativity has been argued to be an important indicator of successful aging. Indeed, generative concern
has been found to be associated with meaning in life. In the present study, this relationship is argued to
be partly explained through generative goals. Moreover, the path between generative goals and meaning
in life is hypothesized to be conditional on belief in the species (assessed through Machiavellianism as
a proxy variable). This moderated mediation model is tested with data from 4 cultural samples: 856
Cameroonian, Czech, German, and Hong Kong Chinese participants aged at least 60 years provided
information on their generative concern, generative goals, meaning in life, and Machiavellianism.
Controlling for effects of relationship, level of education, everyday competence, and cognitive function-
ing on meaning in life, analyses confirmed the moderated mediation model in all cultural samples. That
is, generative concern is partly associated with meaning in life because it leads to generative goals. In
turn, these provide individuals with meaning in life. This association, however, depends on belief in the
species in that meaning in life does not increase when Machiavellian attitudes compete with generative
goals.
Keywords: generativity, belief in the species, goals, well-being, Machiavellianism
Rowe and Kahn (1998) proposed a definition of successful
aging that emphasizes three components: minimizing the risk of
disease and disability, maintaining physical and mental function,
and continuing engagement with (social) life. In our study, we
focus on the last-mentioned facet which Rowe and Kahn argue
consists of cultivation of social relationships as well as productiv-
ity. We do so by stressing the significance of generative strivings
for well-being in old age as generativity offers the opportunity to
be productive (e.g., by offering advise) in a social context (i.e.,
support for the recipient of one’s generative endeavors). Up to
now, few studies have examined the impact of generativity on
human development in late adulthood. Thus, Kruse and Wahl
(2009) identify generativity as a developmental issue that has yet
to be fully acknowledged for its contributions to successful aging.
Given that most findings on determinants and consequences of
generativity stem from Western samples (Arnett, 2008), the gen-
eralizability of what is known about generativity is problematic.
Thus, the present study examines the relationship between genera-
tivity and well-being among elderly people in highly diverse
cultural contexts: Cameroon, the Czech Republic, Germany, and
Hong Kong.
In Erikson’s life span theory, generativity is the central devel-
opmental issue of middle adulthood. Erikson defines generativity
as “the concern in establishing and guiding the next generation”
(Erikson, 1963, p. 276). That is, generative individuals intend to
pass on knowledge and experiences to their junior ones to help
them develop and thrive. Generativity can be expressed in a variety
of social roles (MacDermid, Franz, & De Reus, 1998) and has been
found to affect, for example, parenting behavior (e.g., Peterson,
This article was published Online First November 3, 2014.
Jan Hofer and Holger Busch, Department of Psychology, University of
Trier; Alma Au, Department of Applied Social Sciences, Hong Kong
Polytechnic University; Iva Polácˇková Šolcová, Institute of Psychology,
The Academy of Sciences of the Czech Republic, Prague, Czech Republic;
Peter Tavel, Department of Psychology, Palacky University, Olomouc,
Czech Republic; Teresa Tsien Wong, Department of Applied Social Sci-
ences, Hong Kong Polytechnic University.
This research was supported by a Grant of the German Research Foun-
dation (HO2435/5-1).
For help in recruitment of participants we thank agencies in Hong Kong
(Hong Kong Society for the Aged; Hong Kong Association of Senior
Citizens; Tsim Sha Tsui District Kaifong Welfare Association-
TSTDKFWA Activities Centre for the Elderly; Hong Kong Christian
Service; Kwai Tsing Safe Community and Healthy City Association-Tsing
Yi Community Health Centre; S.K.H. Holy Carpenter Church Community
Centre) and Prague (REMEDIUM, o.p.s.). With respect to the translation
of Chinese questionnaires we thank Professors Helene Fung and Cheng
Sheung Tak.
Correspondence concerning this article should be addressed to Jan
Hofer, University of Trier, Department of Psychology, Developmental
Psychology, D-54296 Trier, Germany. E-mail: hofer@uni-trier.de
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Psychology and Aging © 2014 American Psychological Association
2014, Vol. 29, No. 4, 764–775 0882-7974/14/$12.00 http://dx.doi.org/10.1037/a0037762
764
2006) and political activities (e.g., Peterson, Smirles, & Went-
worth, 1997).
The fact that the developmental crisis of generativity is postu-
lated to be set during middle adulthood does not mean, however,
that generativity is restricted to this life stage (see e.g., Busch &
Hofer, 2011, for generative concern in adolescents). Indeed, Kotre
(1996) argued that various behavioral expressions of generativity
might peak at different points in the life cycle. Specifically, cul-
tural generativity, defined as the passing on and keeping of cul-
turally relevant symbol systems (i.e., ideas, values) is postulated to
peak later in life. Similarly, it has been argued that establishing the next
generation and guiding the next generation might best be concep-
tualized as distinct life stages, with generativity as the establish-
ment of younger generations preceding generativity as the guid-
ance of younger generations (e.g., Vaillant & Milofsky, 1980). In
fact, in his later theorizing, Erikson himself acknowledged that
generativity is of great importance in the process of aging (Erik-
son, Erikson, & Kivnick, 1986): Generative demands may change
with old age, but they do not disappear, and generativity is a vital
element of the elderly’s life review.
In line with this argument, generativity can be regarded as
“evidence of a successful aging process” (Schoklitsch & Baumann,
2011, p. 32). For example, generative people are likely to come to
a positive balance in a life-review, that is, achieve ego-integrity
(James & Zarrett, 2005). Also, asked about their definitions of
successful aging, older people phrase a large share of their re-
sponses in terms of generativity (Fisher, 1995). Generally, aging is
associated with a motivational shift away from goals and projects
that will bring about some personal benefits in the future. Rather,
older people focus on immediate emotional benefits (Carstensen,
Isaacowitz, & Charles, 1999) and on ego-transcending goals
(Brandtstädter, Rothermund, Kranz, & Kühn, 2010;Tornstam,
1997). In line with this reasoning, generativity has been found to
be a goal that is highly esteemed by people who perceive their
future time as limited (as tends to be the case with older individ-
uals; Lang & Carstensen, 2002). Moreover, several studies report
that generative goals become more important with age (e.g., Mc-
Adams, de St. Aubin, & Logan, 1993;Sheldon & Kasser, 2001).
Hence, generativity is an important developmental task in the
aging process.
In an effort to systematize research on generativity, McAdams
and de St. Aubin (1992) have introduced their generativity model.
It consists of seven interrelated aspects: an (a) inner desire and (b)
cultural demands are sources of (c) generative concern, that is, a
positive attitude toward generativity. Generative concern in turn is
a source of (d) generative goals, provided a (e) belief in the species
is present. Generative goals, in turn, are put into practice in (f)
generative action which then is integrated into a person’s (g) life
story in a meaningful way (see McAdams, Hart, & Maruna, 1998,
for more details).
Although the model has been successful in initiating research on
generativity and a variety of predictions have been confirmed
empirically (e.g., Hofer, Busch, Chasiotis, Kärtner, & Campos,
2008; see McAdams et al., 1998), the role of belief in the species
has yet to be examined in detail. Erikson (1963) argues that people
may show behaviors that have a generative appearance but still fail
to be truly generative. For example, parenthood may be seen as a
prototypically generative behavior (cf. Kotre’s, 1996, concepts of
biological and parental generativity). Still, Erikson (1963) says, to
be a truly generative parent you need the conviction that your
generative effort is worthwhile. That is, you have to be convinced
that humankind develops in a positive way and that those intended
to receive your generative efforts will appreciate this contribution
(e.g., McAdams, 2000;McAdams & de St. Aubin, 1992). Other-
wise, one would be forced to question the worth of one’s gener-
ative behavior: To use Erikson’s example, why have children if
humankind turns the world into a worse place to live in every day?
Why teach my children about what I have learnt in life if he or she
will dismiss what I am saying anyway?
In their generativity model, McAdams and de St. Aubin (1992)
take up Erikson’s idea of belief in the species by proposing that
without this belief people find it hard to commit to generative
goals. More specifically, they postulate that generative goals are
stimulated by generative concern. Generative concern is the first
conscious representation of generativity in the individual: People
care for how the following generations fare. That is, generative
concern denotes a general willingness to help younger generations
thrive. This general willingness to become generative is then
translated into more specific generative goals (for empirical con-
firmations of this relation see Hofer et al., 2008;McAdams et al.,
1993). According to the generativity model, however, this rela-
tionship is qualified by belief in the species: When belief in the
species is present, generative concern is easily translated into
generative goals; its absence, on the other hand, reduces the extent
to which generative concern results in generative goals.
Again, it is unfortunate that belief in the species has hardly
received any attention in research on generativity. Thus, the exact
role of belief in the species has not yet been examined. A major
reason for this lack of research is that no instrument has been
established that specifically assesses belief in the species yet.
Instead, researchers have tried to approach the belief in the species
concept by proxy variables. For example, Van De Water and
McAdams (1989) employed scales such as, for example, faith in
people and basic trust to tap belief in the species. This attempt,
however, has not yet received any replication.
More recently, Machiavellianism has been argued to be a proxy
for belief in the species (Busch & Hofer, 2012). Machiavellianism
is a personality trait that combines a cold and manipulative view
on others and a cynical moral stance so that behavior is selected
which allows a maximal satisfaction of the Machiavellian’s inten-
tions (for an overview, see Jones & Paulhus, 2009). Consequently,
Machiavellians have frequently been found to lie or cheat (Jones &
Paulhus, 2009). That is, they readily manipulate others to achieve
their goals and disprize other people’s concerns because they see
others as means rather than ends (Christie & Geis, 1970). McHo-
skey (1999) demonstrated that people high in Machiavellianism
have little interest in contributing to society. Moreover, if they do
show prosocial behavior, Bereczkei, Birkas, and Kerekes (2010)
argue, it needs to be public so they gain prestige and recognition.
Thus, ultimately, even seemingly prosocial behavior serves a self-
ish agenda. In sum, the cynical and self-centered Machiavellian
perspective seems hardly compatible with the “hope in the ad-
vancement and betterment of human life in succeeding genera-
tions” characteristic of belief in the species (McAdams & de St.
Aubin, 1992, p. 1006). Thus, Machiavellianism is well suited to
capture the low end of belief in the species.
Although McAdams and de St. Aubin’s (1992) model of gen-
erativity assumes that belief in the species affects generative goal
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
765
GENERATIVITY IN OLD AGE
commitment, this effect has not yet been subject to empirical
examination. In fact, the first study on belief in the species found
the proxy variables representing belief in the species to be asso-
ciated with self-reported generative behavior arguing that belief in
the species is “a key prerequisite for generative action” (Van De
Water & McAdams, 1989, p. 437). This suggests that it might not
only be the setting of generative goals which is affected by belief
in the species but the actualization of generative goals as well:
Generative goals may be set in spite of low belief in the species,
but under such circumstances generative goals fail to be actualized
in corresponding behavior. Thus, a lack of belief in the species
might affect consequences of generative goal setting as much as
generative goal setting per se. We propose that people lacking
belief in the species would not derive any meaning from their
generative goals.
Generally, goals provide individuals’ lives with meaning and a
sense of purpose (e.g., Bühler & Massarik, 1968;Frazier, New-
man, & Jaccard, 2007). Indeed, Ryff (1989) proposes that having
goals is an aspect of positive functioning in its own right. As
Emmons (2003, p. 107) puts it, “goals are essential components of
a person’s experience of his or her life as meaningful and contrib-
ute to the process by which people construe their lives as mean-
ingful and worthwhile” (note that the latter thought is reflected in
the fact that generative individuals tend to emphasize prosocial
goal pursuit in their life stories; McAdams, Diamond, de St.
Aubin, & Mansfield, 1997). Goal pursuit is thus reliably found to
contribute to well-being (e.g., Diener, 1984;Emmons, 1986).
However, the full potential of goal pursuit to foster well-being can
be amplified by internal factors as for example in the case of goals
that are intrinsic rather than extrinsic (see Deci & Ryan, 2000, for
an overview) or congruent with one’s implicit motive system (e.g.,
Hofer, Busch, Bond, Li, & Law, 2010). We assume that in the
context of generativity, belief in the species is such a factor. Thus,
the concurrence of generative goals and strong belief in the species
ought to be more beneficial for well-being than generative goals
that are pursued despite a low belief in the species.
Previous research, however, has focused on the relation of
generative concern rather than generative goals with well-being.
Generative concern has consistently been found to promote life
satisfaction (e.g., Ackerman, Zuroff, & Moskowitz, 2000;Hofer et
al., 2008) and purpose in life (e.g., Busch & Hofer, 2012;Gross-
baum & Bates, 2002). That is, generative concern has already been
demonstrated to be associated with meaning in life. To the best of
our knowledge, however, no study has considered the relation of
generative concern and generative goals with meaning in life
simultaneously.
The Present Research
To sum up, for reasons given above it is hypothesized that the
relationship between generative concern and meaning in life is—at
least—partly mediated by generative goals. Hence, we assume that
generative concern as a positive attitude toward generativity mo-
tivates people to set corresponding goals which then provide them
with a sense of meaning in life. Yet, the mediation effect is
hypothesized to be moderated by belief in the species. Only high
levels of belief in the species will allow elderly people to perceive
enhanced levels of meaning in life by pursuing generative goals.
Due to the lack of a direct measure of belief in the species,
Machiavellianism will be used as an indicator of a lack of belief in
the species (see Busch & Hofer, 2012). In Figure 1, the hypothe-
sized relationships among psychological constructs are presented.
Additionally, measurements of physical and cognitive competence
were included to control for effects on elderly people’s well-being
(Waddell & Jacobs-Lawson, 2010).
A second focus derives from the cross-cultural design. The
present study assumes that the psychological mechanisms delin-
eated above can be generalized to a variety of cultural contexts. It
is argued that culture-bound patterns of socialization affect indi-
viduals’ cognitive, emotional, and motivational processes (Keller,
2007;Markus & Kitayama, 1991). Thus, the cultural context
shapes developmental processes and affects an individual’s behav-
ior and strivings by representing a major source of (behavioral)
determinants that have their locus outside the individual (Pepitone,
1976). Yet, there is latitude for variation in individuals’ basic
components of personality (e.g., motives, traits) among members
of a given cultural group arising from biological and educational
variations. Additionally, recent research has shown that despite a
surface diversity in cultural norms and values, there are certain
invariant aspects of basic human psychological processes (e.g.,
Hofer et al., 2010; for aging-related findings see Fung, Carstensen,
& Lutz, 1999). Thus, we challenge the idea that culture per se
results in variability in basic psychological mechanisms or pro-
cesses and hypothesize that the proposed relationship between
psychological constructs is found across diverse cultural contexts.
To make such a cross-cultural generalization as broad as possi-
ble, it is indispensable to assess data from cultural samples repre-
senting a wide range of cultural marker variables (van de Vijver &
Leung, 1997). The selection of cultures in the present study was
based on value orientations (Schwartz, 1992), which show consid-
erable overlap with other cultural markers such as individualism/
collectivism (e.g., Triandis, 1996). Thus, participants were re-
cruited in Cameroon, China (Hong Kong), the Czech Republic,
and Germany because recent research has provided evidence that
Cameroonian, Chinese, Czech, and German participants signifi-
cantly differ from each other in their value orientations (e.g., Hofer
et al., 2010;Jowell & the Central Coordinating Team, 2007). Still,
this assumption guiding the selection of cultures for the present
study was tested by including a measure of value orientations.
Method
Sample
To assure cultural homogeneity within the four samples, only
native participants were recruited. With respect to the multiethnic
Figure 1. Model on mediational effect of generative goals on the rela-
tionship between generative concern and meaning in life moderated by
Machiavellianism.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
766 HOFER ET AL.
population in Cameroon, sampling was restricted to ethnic
Grassfield Bantus (Nso) from the Anglophone northwest prov-
ince in Cameroon (Mbaku, 2005). Chinese participants were
recruited in Hong Kong and recruitment of Czech participants
was done in Prague and Olomouc. Finally, German participants
were recruited in Osnabrück.
In total, data for present analyses were collected from 856
noninstitutionalized participants (for sociodemographic informa-
tion see Table 1). Of these, 225 were from Cameroon, 243 from
China, 163 from the Czech Republic, and 224 from Germany.
With exception of the Chinese sample, cultural samples were
reasonably balanced concerning gender. Participants ranged in age
from 60 to 90 years. Women and men did not differ in mean age.
Yet, cultural samples differed in mean age (F
(3, 852)
⫽78.14; p⬍
.001;
2
⫽.22): Chinese and Czech participants were significantly
older than Cameroonian and German participants who in turn also
differed significantly from each other in age. Eta-squared (
2
)is
reported as index of the strength of association;
2
s of .01, .06, and
.14 can be interpreted as small, medium, and large effect size,
respectively.
Participants’ level of education was categorized into three cat-
egories, that is, low (primary school education or less), medium
(junior secondary school or ordinary level education), and high
education (secondary school or university education). Analyses
revealed differences between cultural samples in the distribution of
educational levels (
2(6)
⫽296.49; p⬍.001): Low levels of
education were more often assigned to Cameroonian and Chinese
participants. Additionally, a medium level of education could not
be assigned to Czech participants due to types of degrees in the
educational system.
In total, 560 participants were in a current relationship (married:
n⫽540) and 296 participants (widowed: n⫽192) reported to live
without a steady partner. Although the percentage of participants
having a steady relationship did not greatly differ between samples
from China, the Czech Republic, and Germany, it was highest
among Cameroonian participants (
2(3)
⫽22.46; p⬍.001).
Only 61 of the study participants were childless, and 795 had at
least one child (range: 1 to 18; M⫽3.23; SD ⫽2.45). Analyses
showed (F
(3, 852)
⫽220.91; p⬍.001;
2
⫽.44) that the highest
number of children born was reported by Cameroonian partici-
pants, followed by Chinese participants, and, finally, by Czech and
German participants. The latter two did not differ from each other
in number of children.
Procedure
Recruitment in Germany was done via local newspapers. In the
Czech Republic and China, elderly people were contacted via
flyers and notes in local senior centers. Typically, elderly people in
Cameroon return to their home village after retirement. Thus, local
research assistants visited villages near major cities in the north
province of Cameroon to recruit elderly people for the present
study.
All participants were volunteers and were guaranteed that any
information given would be treated confidentially. Cameroonian
and German participants were given monetary compensation pro-
portional to average differences in GDP per capita. Chinese and
Czech study subjects received coupons to buy goods at will in
local supermarkets. Data collection was done on university prem-
ises in Germany and in local senior centers in Hong Kong; in the
other regions, it was mostly conducted at the participants’ homes.
At all research sites, local research assistants familiar with all
instruments applied were present while participants worked
through questionnaires to help clarify any questions that arose.
Instruments were administered in either Chinese (China), Czech
(the Czech Republic), or German (Germany). In Cameroon Eng-
Table 1
Descriptive Statistics of Sociodemographic Information, Value Orientations, Meaning in Life, Generative Concern, Generative Goals,
Machiavellianism, Instrumental Activities of Daily Living, and Cognitive Competence in the Four Cultural Samples
Measurement Cameroon
M(SD)China
M(SD)Czech Rep.
M(SD)Germany
M(SD)
Sociodemographic information
1. Age 64.60 (5.62) 72.63 (7.41) 71.45 (5.55) 67.49 (5.78)
2. Gender (% females) 48.7 72.8 56.4 54.9
3. Education (% low/medium/high) 63.3/16.8/19.9 52.7/25.9/21.4 9.8/0/90.2 21.0/30.4/48.7
4. Relationship status (% without partner) 22.6 39.1 43.6 35.3
5. Number of children born 5.88 (2.79) 2.78 (1.61) 1.96 (.93) 1.98 (1.30)
Value orientations (selection of cultural groups)
6. Openness to change 3.06
c
(1.34) 3.51
b
(1.21) 2.99
c
(.99) 4.20
a
(1.06)
7. Conservation 5.21
a
(.59) 5.28
a
(.91) 4.52
b
(.89) 3.98
c
(1.02)
Psychological variables for structure-oriented analyses
8. Meaning in life 4.38 (.76) 4.10 (1.13) 4.36 (1.00) 4.25 (1.03)
9. Generative concern 2.06 (.44) 1.11 (.61) 1.52 (.49) 1.46 (.52)
10. Generative goals 3.35 (.50) 3.12 (.73) 3.32 (.59) 3.03 (.64)
11. Machiavellianism 1.30 (.49) 1.37 (.61) 1.70 (.64) 1.29 (.62)
Everyday and cognitive competence (control variables in structure-oriented analyses)
12. Everyday competence 2.57 (.65) 3.11 (.55) 3.06 (.47) 3.34 (.38)
13. Cognitive competence
1
.54 (1.31) .27 (.83) ⫺.47 (.66) ⫺.50 (.48)
Note.
a, b, c
⫽Different subscripts for value orientations indicate statistically significant differences between cultural samples. See text for details.
1
⫽
Descriptives for the subtests of the NAI within the four cultural samples: ZVT-G (CAM: M⫽58.05; SD ⫽37.19; CH: M ⫽40.07; SD ⫽21.35; CZ: M⫽
29.54; SD ⫽10.85; GER: M⫽25.14; SD ⫽8.68); LT-G (CAM: M⫽59.30; SD ⫽38.73; CH: M⫽49.88; SD ⫽24.45; CZ: M⫽36.94; SD ⫽25.69;
GER: M⫽35.06; SD ⫽16.78); BT (CAM: M⫽5.15; SD ⫽1.01; CH: M⫽4.18; SD ⫽1.22; CZ: M⫽5.72; SD ⫽1.06; GER: M⫽5.33; SD ⫽1.02).
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
767
GENERATIVITY IN OLD AGE
lish versions of instruments were administered. Although English
is not the native language among the Nso in the northwest province
of Cameroon, it represents the official language and is exclusively
used in educational institutions and predominantly so in everyday
life. Moreover, only very few people are able to read or write in the
colloquial languages in Cameroon. However, local assistants who
were all ethnic Nso were trained to give (standardized) illustrations
of test items in Lamnso (native language of the Nso) or Pidgin-
English when necessary.
Measurements
Measurements applied for the present study were part of a
cross-cultural project on successful aging in different cultural
contexts. Instruments were administered to participants individu-
ally. First, the questionnaires on meaning in life, life goals, value
orientations, generative concern, Machiavellianism, and activities
of daily living (ADLs) were given. Next, participants provided
information on their sociodemographic characteristics and took a
brief test on their cognitive performance.
Although some instruments were available in all language ver-
sions needed (e.g., the Schwartz Value Survey, SVS; Schwartz,
1992), others had to be translated into Chinese and/or Czech. The
lacking Chinese instruments were translated from their original
English versions by bilingual research assistants in Hong Kong.
The lacking Czech instruments were translated from their English
or German version by a professional translator into Czech. The
quality of all translated material was ensured by a back-translation
by bilingual psychologists.
Meaning in life. Perceived meaning in life was assessed with
the subscale presence of meaning (MLQ-P; e.g., “My life has a
clear sense of purpose”) of the meaning in life questionnaire
(Steger, Frazier, Oishi, & Kaler, 2006). Participants are asked to
rate each of the five items on a 7-point Likert scale ranging from
0(strongly disagree)to6(strongly agree). Cronbach’s Alpha for
the total sample was .79 (ranging from .72 among Cameroonian to
.83 among Chinese participants).
Generative goals. The questionnaire GOALS (Pöhlmann &
Brunstein, 1997) measures the importance participants assign to
goals from various life domains. Participants are asked to rate
goals on a 5-point Likert scale from 0 (not at all important)to4
(very important). In the present study, four items measuring the
importance of altruistic goals (i.e., support other people in their
efforts; act selflessly; do good; help other people who are in need)
were used to assess generative strivings of elderly people. Cron-
bach’s Alpha was .77 for the total sample (ranging from .71 in the
Cameroonian to .79 in the Chinese sample).
Value orientations. As a cultural marker for differences be-
tween samples, the Schwartz Value Survey (Schwartz, 1992) was
administered to assess the importance that individuals ascribe to
certain value orientations as guiding principles for their lives. In
the present study, only 22 items were used that have universally
been shown to have the same meaning with respect to the higher
order value types conservation (14 items; subsumes the value
types of conformity, security, and tradition; e.g., honoring parents
and elders; family security; humble) and openness to change (eight
items; subsumes the value types of self-direction and stimulation;
e.g., choosing own goals; varied life). These two value dimensions
were selected as openness to change with its emphasis on auton-
omy and self-direction relates to individualism, and conservation
with its emphasis on self-restriction and protection of stability to
collectivism (Triandis, 1996). Items are rated on a Likert scale
from ⫺1(opposed to my values)to7(of supreme importance). In
total, datasets of 834 (openness to change) and 821 (conservation)
participants, respectively, could be used for analyses. Cronbach’s
Alpha was .86 for conservation (ranging from .75 in the Camer-
oonian to .85 in the German sample) and .79 for openness to
change (ranging from .74 in the Czech to .81 in the German and
Cameroonian samples).
Generative concern. The Loyola Generativity Scale (LGS;
McAdams & de St. Aubin, 1992) was administered to assess
generative concern. The LGS includes 20 items (e.g., “I try to pass
along the knowledge I have gained through my experience”)
designed to measure a general disposition for generativity. Partic-
ipants are asked to evaluate items on a 4-point Likert scale from 0
(never)to3(very often). Cronbach’s Alpha was .86 (ranging from
.75 in the Cameroonian to .84 in the Czech sample).
Machiavellianism. The tendency to view people as objects to
be manipulated and exploited with emotional detachment and cool
calculation was measured by administering the Mach IV (Christie
& Geis, 1970). In the present study only the 10 positively phrased
statements reflecting Machiavellian views of human nature (e.g.,
“Anyone who completely trusts anyone else is asking for trouble”)
were used. The items are evaluated on a 5-point Likert scale
ranging from 0 (strongly disagree)to4(strongly agree). Cron-
bach’s Alpha was .66 (ranging from .60 among Cameroonian to
.75 among German participants).
Everyday competence. Effective coping with daily demands
was assessed by reports on difficulties in instrumental activities of
daily living (IADL; Lawton & Brody, 1969). Originally, 16 items
covering various domains of activities (e.g., food preparation and
shopping; handling of finances and medication; housekeeping;
modes of transport) were used. However, initial analyses showed
that two items (e.g., daily routines in meals) did not add to the
reliability of the construct in the Czech and German samples.
Thus, the final IADL-scale included 14 items. Participants rated
each item (e.g., “In case I have to take some medicine, I always
know exactly when and at which dosage to take it”) on a 5-point
Likert scale ranging from 0 (strongly disagree)to4(strongly
agree). Cronbach’s Alpha was .85 (ranging from .73 in the Ger-
man to .84 in the Cameroonian and Chinese samples).
Cognitive performance. In contrast to self-reported difficul-
ties in IADLs, three subtests of the Nürnberger Altersinventar
(NAI; Oswald & Fleischmann, 1997) were administered to assess
participants’ cognitive speed and memory performance. Cognitive
functioning has been described as an indicator of successful aging
(e.g., Rowe & Kahn, 1998) and relates to measurements of sub-
jective well-being among elderly people (e.g., Staudinger &
Fleeson, 1996). In the present study, two subtests for speed of
information processing were administered, that is, the “number-
connection test” (ZVT-G; two exercise and two test sheets) and the
“labyrinth test” (LT-G; one exercise and one test sheet): (a) On
each test sheet of the ZVT-G, the numbers 1 to 30 are printed in
a scrambled order and have to be connected with a pen as fast as
possible. Mean time in seconds for the two test trials indicates
cognitive speed; (b) The LT-G measures efficiency of visuomotor
coordination: By using a pen, participants are asked to find the way
out of a labyrinth as fast as possible. The time of the test trial is the
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
768 HOFER ET AL.
indicator of cognitive functioning. Finally, memory performance
was assessed by a third subtest, the “picture-test” (BT). Partici-
pants are shown seven picture cards (each for approximately 3 s;
e.g., screwdriver, button). Immediately after the presentation, par-
ticipants are asked to recall the pictures. The number of correct
pictures represents the index for memory performance.
The three criteria for cognitive speed and memory functioning
are combined to form one single indicator of cognitive perfor-
mance. Intercorrelations among the measures were sufficiently
high with r⫽.72 between the two speed tests (p⬍.001) and
r⫽⫺.25 (LT-G) and ⫺.23 (ZVT-G; p⬍.01), respectively,
between the memory performance and speed tests. To produce a
single measure of cognitive performance the three test scores were
combined based on the regression factor scores for the first com-
ponent derived from principal components analysis that accounted
for 62% of the variance within the total sample. Factor loadings of
the emerged one-factor solution were .90 (LT-G), .89 (ZVT-G),
and ⫺.50 (BT). Thus, low scores indicate good cognitive perfor-
mance.
Equivalence of measurements across cultural groups. With
the exception of differences in value orientations (cultural marker),
the present study has a structure-oriented rather than a level-
oriented focus. That is, the emphasis is on the relationship among
psychological constructs across cultural groups rather than differ-
ences in mean levels of test scores between cultural samples.
However, data on psychological constructs included in the hypoth-
esized model (generative concern, generative goals, meaning in
life, Machiavellianism) were examined for measurement invari-
ance across cultural samples to guarantee that (latent) variables in
the hypothesized models are psychometrically sound.
1
Internal consistencies indicate that sets of items can be treated as
measuring single latent variables across cultural samples under
investigation. Additionally, none of the scales showed an internal
consistency that falls below an acceptable ␣of .6 (Nunnally,
1978). To further scrutinize cross-cultural applicability of data,
exploratory factor analyses (EFA; principle component analysis)
were conducted separately for each of the four scales within each
cultural sample. With respect to measurements of generative goals,
meaning in life, and Machiavellianism, findings indicate that all
items show significant factor loadings on a single factor within
each cultural sample. Ranges of factor loadings across the four
samples are .60 to .87 for generative goals, .47 to .86 for meaning
in life, and .27 to .71 for Machiavellianism. Results of the EFA for
generative concern were less straightforward. Analyses indicated
that all six reverse coded items of the LGS should be excluded
from further analyses because of nonsignificant factor loadings in
at least one cultural group. Additionally, the item referring to
adopting children was excluded as it did not show a significant
factor loading in the Cameroonian and Czech data.
Next, measurement equivalence of generative goals, meaning in
life, Machiavellianism, and the reduced scale for generative con-
cern (13 items) across the cultural subsamples was examined in
more detail by use of multigroup confirmatory factor analysis. It
was tested whether the measurement model upholds by specifying
two models, one including the data on meaning in life and Ma-
chiavellianism, and the other the data on generative goals and
generative concern.
According to Kline (1998), the ratio of cases/observations to
number of free parameters ought to be in the range of 10 or even
20. Thus, to reduce the number of free parameters to be estimated
in the measurement models, item parcels were built for the mea-
surements of Machiavellianism and generative concern. Although
item parceling is controversially discussed in literature (e.g., Little,
Cunningham, Shahar, & Widaman, 2002), it is an appropriate
means in SEM procedures if unidimensionality of constructs under
investigation is given (e.g., Bandalos, 2002). EFAs approved uni-
dimensionality of measurements. For building parcels, we used
random assignment resulting in homogenous samples that were
similar in variance. With respect to Machiavellianism, three par-
cels were built that included three or four items. Similarly, the
three parcels built for the remaining item set of the LGS contained
four or five items.
The first measurement model (Model A) included the two latent
variables meaning in life and Machiavellianism which were mea-
sured by five items and three parcels, respectively. The second
measurement model to be tested (Model B) included the latent
variables generative goals (four items) and generative concern
(three parcels). In the specified models latent factors were allowed
to be correlated. Two increasingly restrictive measurement models
were defined: first, the unconstrained model with no equality
constraints across cultural groups; second, the measurement
weights model where the measurement weights were constrained
to be equal across all groups and the variances and covariance of
the two latent scores were estimated separately for each group.
Referring to Model A (meaning in life and Machiavellianism),
the unconstrained model (144 data points, 68 unknown parame-
ters) adequately fit our data (
2
/degrees of freedom: 1.77; CFI: .97;
RMSEA: .030). Within each of the cultural samples, all items/
parcels significantly loaded on the specified factor with critical
ratios (CR) ⱖ3.03 (p⬍.01). Also, the measurement weights
model with factor loadings constrained to be equal across cultural
samples showed a good data fit (
2
/degrees of freedom: 1.91; CFI:
.96; RMSEA: .033).
With respect to Model B (generative goals and concern), both,
the unconstrained model (112 data points; 60 unknown parame-
ters;
2
/degrees of freedom: 1.81; CFI: .98; RMSEA: .031; CRs ⱖ
5.10; p⬍.001) and the measurement weights model (
2
/degrees
of freedom: 1.93; CFI: .97; RMSEA: .034) adequately fit the data.
Using chi-square difference tests, nested model comparisons
indicated an impairment of fit when constraining factor loadings
(Model A:
2
(18) ⫽45.08; Model B:
2
(15) ⫽35.44). However,
the chi-square statistic is dependent on sample size and small
differences become easily significant if sample size is large. Thus,
evaluating the fit of two nested models ought to be done from the
perspective of multiple fit indices (Cheung & Rensvold, 2002).
According to recommendations set forth by Chen (2007) and
Cheung and Rensvold (2002), the hypotheses of measurement
invariance should not be rejected when change in fit indices from
1
As the SVS has already been widely and successfully used in more
than 80 cultural contexts around the globe (e.g., Schwartz, 1992) and data
on value orientations are only used as cultural markers we could refrain
from testing measurement invariance. Likewise, data on IADLs which
served to control for everyday competence in individuals’ meaning in life
were excluded from detailed analyses on measurement invariance. Yet,
internal consistencies of value dimensions and everyday competence, re-
spectively, point to applicability of measurements in the cultural samples at
hand.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
769
GENERATIVITY IN OLD AGE
the unconstrained to the more constrained model is low (e.g.,
⌬CFI ⱕ⫺.01; ⌬RMSEA ⱕ.015). Following these guidelines,
constraining factor loadings across cultural samples did not result
in an impairment of fit (Model A: ⌬CFI ⫽⫺.01; ⌬RMSEA ⱕ
.003; Model B: ⌬CFI ⫽⫺.01; ⌬RMSEA ⱕ.003).
To conclude, analyses on measurement equivalence indicate
configural and metric invariance allowing us to meaningfully
examine structural relationships among psychological constructs
across cultural groups. In subsequent analyses, the reduced gener-
ative concern scale (13 items) will be used. Cronbach’s Alpha was
.86 (ranging from .81 in the Cameroonian to .86 in the Chinese
sample).
Results
Results will be presented according to the following schema:
First, to scrutinize whether cultural samples were appropriately
selected, assumptions on cultural orientations in the present sam-
ples will be tested by examining mean differences in value orien-
tations. Next, correlations among psychological measurements in
the total sample as well as correlations between psychological
measurements and individuals’ sociodemographic characteristics
and their everyday competence and cognitive functioning will be
given. The latter set of correlations is conducted to identify effects
of sociodemographic characteristics as well as everyday compe-
tence and cognitive functioning on the dependent variable, that is,
meaning in life. Variables impacting meaning in life will be
controlled for in subsequent analyses. Finally, analyses on the fit
of the mediation model (generative concern, generative goals, and
meaning in life) and the moderated mediation model (moderating
effect of Machiavellianism) for the total sample but also across
cultural groups will be presented. Descriptive data on measures are
presented in Table 1.
Differences in Value Orientations
Between Cultural Samples
A multivariate analysis of variance (MANOVA) with higher-
order value types as dependent variables and culture as factor was
conducted to test for differences in higher order value types
between cultural samples. Multivariate statistics indicated variabil-
ity across cultural groups (F
(6, 1610)
⫽99.37/Wilks’s criterion; p⬍
.001;
2
⫽.97). Culture showed a significant effect on openness
to change (F
(3, 806)
⫽42.72; p⬍.001;
2
⫽.14) and conservation
(F
(3, 806)
⫽105.98; p⬍.001;
2
⫽.28). Post hoc tests (Bonfer-
roni) showed that German participants scored significantly higher
for openness to change than Czech, Chinese, and Cameroonian
participants (ps ⬍.001). Additionally, Chinese participants re-
ported higher levels of openness to change than Cameroonian and
Czech participants (ps ⬍.001). With respect to conservation,
German participants scored lowest (ps ⬍.001). Finally, Czech
participants showed lower levels of conservation than Cameroo-
nian and Chinese participants (ps⬍.001). To conclude, findings
indicate significant variation across important dimensions of value
orientation that are typically used as cultural markers, and sub-
stantiate the assumptions guiding the selection of cultural groups
for the present study.
Correlations Among Psychological Measurements as
Well as Between Psychological Measurements and
Sociodemographic Variables, Everyday Competence,
and Cognitive Functioning
In the upper part of Table 2, correlations among psychological
measurements are given. As shown, meaning in life was signifi-
cantly positively associated with generative concern and goals, and
negatively related to Machiavellianism. Furthermore, generative
concern showed a significant positive association with generative
goals. Level of Machiavellianism was unrelated to measurements
of generative concerns and goals.
2
In the lower part of Table 2, correlations between psychological
measurements and participants’ age, gender, level of education,
relationship status, number of children, everyday competence, and
cognitive functioning are presented. With respect to meaning in
life, that is, the dependent variable in subsequent analyses, higher
levels of meaning in life were significantly associated with both a
higher degree of school education and having a steady relationship.
Although no significant effects could be found for participants’
age, gender, and number of children, everyday and cognitive
competence were significantly related to meaning in life: The
better participants’ everyday and cognitive competence, the higher
were the reported levels of meaning in life. Thus, participants’
level of education and relationship status as well as their everyday
and cognitive competence were controlled for in subsequent anal-
yses on meaning in life.
Significant associations were also identified for generative con-
cern: Age and number of children were significantly positively
related to generative concern. Furthermore, higher reports of gen-
erative concern were significantly related to a higher level of
education, to having a steady partner, and to being male, but not to
measures of everyday and cognitive competence. Only a higher
number of children born was significantly associated with gener-
ative goals. Finally, higher levels of Machiavellianism were sig-
nificantly associated with a higher age and with being male.
Testing the Mediation
and Moderated Mediation Model
In the final part of the result section, analyses on the proposed
models of mediation and moderated mediation will be presented.
First, findings derived from the total sample will be given. Next,
the fit of the hypothesized models across cultural groups will be
examined.
Testing the mediation effect of generative goals. We hy-
pothesized that commitment to generative goals mediates the link
between generative concern and meaning in life. The simple me-
diation, which can be considered to represent an average model,
was tested by applying Preacher and Hayes’ (2004) model for
estimating the size and significance of the indirect effect in simple
mediation. In analyses, the dependent variable, that is, meaning in
life, was controlled for effects of relationship status, level of
education, and everyday and cognitive competence. Criteria for
2
As examining the fit of the mediation and moderated mediation models
across cultural samples involves a testing of differences in strength of paths
between constructs, we could refrain from testing differences in strength of
correlational patterns between cultural samples.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
770 HOFER ET AL.
mediation were met (see Figure 2): Generative concern signifi-
cantly related to meaning in life and generative goals. Addition-
ally, significant indirect effects of commitment to generative goals
could be verified in analyses on meaning in life (indirect effect ⫽
.04; SE ⫽.02; z⫽2.54; p⬍.05; 95% confidence interval for
generative goals based on 1,000 bootstrap resamples: [.01, .08]).
Testing the moderated mediation hypothesis. To examine
whether, and if so, how the effect of the mediator on the dependent
variable varies across levels of the moderator, the integrative
approach suggested by Preacher, Rucker, and Hayes (2007) was
applied. This approach allows combining mediation and modera-
tion into one general analytical framework (see also Edwards &
Lambert, 2007). Thus, in the hypothesized model generative goals
mediate the effect of generative concern on meaning in life.
However, Machiavellianism is assumed to moderate the path from
generative goals to meaning in life. As the distribution of product
terms is usually skewed, conditional indirect effects will be exam-
ined in detail by employing a bootstrap approach.
Coefficient estimates in Figure 3 show again that there was a
significant direct effect of generative concern on meaning in life
and generative goals which in turn were significantly related to
meaning in life. Most important for our hypothesis, findings indi-
cated that the indirect effect of generative concern on meaning in
life via generative goals was moderated by Machiavellianism.
Conditional indirect effects showed that there were significant
indirect effects of generative concern on meaning in life only for
low levels of Machiavellianism, that is, one standard deviation
below the mean (indirect effect ⫽.07, SE ⫽.02, z⫽3.07, p⬍.01;
95% confidence interval for the indirect effect one SD below the
mean based on 1,000 bootstrap resamples: [.03, .12]) and medium
levels of Machiavellianism, that is, at the mean (indirect effect ⫽
.04, SE ⫽.02, z⫽2.25, p⬍.05; 95% confidence interval for the
indirect effect at the mean: [.01, .07]). For high levels of Machi-
avellianism (one standard deviation above the mean) the indirect
effects did not reach significance (p⫽.80).
In Figure 4, the moderating effect of Machiavellianism on the
relationship between generative goals and meaning in life is
graphed. As shown, the relationship between generative goals and
meaning in life was steeper for respondents who reported low and
medium rather than high levels of Machiavellianism. Among par-
ticipants characterized by high levels of Machiavellianism gener-
ative goals were not associated with meaning in life.
3
Testing the fit of the hypothesized models across cultural
groups. Finally, the equivalence of the proposed mediation and
moderation effects across cultural groups was examined by apply-
ing multigroup, structural equation modeling (SEM) that has be-
come a preferred method in multivariate analysis in social and
behavioral sciences when, for example, the strength of relation-
ships among measurements is tested for equivalence in various
(cultural) groups.
Multigroup path analyses with manifest variables were con-
ducted to test equality of regression coefficients across cultural
groups by applying SEM with maximum likelihood estimation
(AMOS). The moderated mediation model depicted in Figure 1
was tested. Generative concern, generative goals, Machiavellian-
ism, and the interaction term of generative goals with Machiavel-
lianism were treated as exogenous variables, whereas meaning in
life was treated as an endogenous variable with effects of level of
education, relationship status, everyday competence, and cognitive
3
Support for our hypothesized model stems from additional analyses
showing that Machiavellianism did not moderate the path from generative
concern to generative goals.
Table 2
Correlations Among Meaning in Life, Generative Concern and Goals, and Machiavellianism and
Between Psychological Measurements and Sociodemographic Variables, Everyday Competence,
and Cognitive Functioning
Measurement 1
Meaning in life 2
Generative concern 3
Generative goals 4
Machiavellianism
1—
2 .31
ⴱⴱⴱ
—
3 .17
ⴱⴱⴱ
.30
ⴱⴱⴱ
—
4⫺.13
ⴱⴱⴱ
⫺.03 ⫺.05 —
Age ⫺.03 .27
ⴱⴱⴱ
⫺.01 .11
ⴱⴱ
Gender
1
.05 .14
ⴱⴱⴱ
⫺.03 .11
ⴱⴱ
Level of Education .12
ⴱⴱ
.08
ⴱ
⫺.02 .03
Relationship status
1
.09
ⴱ
.10
ⴱⴱ
.01 ⫺.03
Number of children born .06 .33
ⴱⴱⴱ
.12
ⴱⴱ
⫺.04
Everyday competence .09
ⴱ
⫺.05 ⫺.06 .05
Cognitive competence ⫺.14
ⴱⴱⴱ
.03 ⫺.01 .02
Note.
1
⫽Gender and relationship status were coded 0 (female; no partner) and 1 (male; with partner).
ⴱ
p⬍.05.
ⴱⴱ
p⬍.01.
ⴱⴱⴱ
p⬍.001.
Figure 2. Generative goals mediate between generative concern and
meaning in life. Note: Coefficients are given as follows: (b/SE).
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
771
GENERATIVITY IN OLD AGE
functioning being partialed out. Two models were defined: first,
the unconstrained model in which no equality constraints were
imposed on the data across cultural groups and exogenous vari-
ables were allowed to correlate with each other, and second, the
structural weights model in which paths were set to be equal in all
four cultural groups. A good fit of this second model would
indicate that the effect of the predictors on the outcome variable,
that is, meaning in life, did not differ between cultural samples.
The two defined models were tested against each other through
nested model comparisons.
After equality constraints on regression coefficients were placed
across groups (intercepts were left free to vary), the constrained
models gained 15 degrees of freedom (60 data points minus 45
parameters to be estimated). Fit indices indicated that the specified
structural weights model (all paths set as equal across cultural
groups) fit the data sufficiently well. The ratio of the chi-square
value degrees of freedom was 1.62. The Adjusted Goodness of Fit
Index (AGFI ⫽.96) exceeded its critical value (.90); the Root
Mean Square Error of Approximation (RMSEA ⫽.027) fell below
its critical value (.06) and the Akaike’s Information Criterion
(AIC ⫽114.31) was lower than the critical value (120.00), as
determined by a saturated model. Most importantly, nested model
comparisons showed that the structural weights models did not fit
worse than the unconstrained models, ⌬
2
(15) ⫽24.31, p⬎.05.
Referring to the structural weights model, all structural parameter
estimates were statistically important. In detail, there was a sig-
nificant relationship between generative concern and generative
goals (CR ⫽7.36; p⬍.001; s ranging from .24 to .25). Fur-
thermore, meaning in life was significantly linked to generative
concern (CR ⫽7.64; p⬍.001; s ranging from .24 to .28), to
generative goals (CR ⫽2.05; p⬍.05; s⫽.07), to Machiavel-
lianism (CR ⫽⫺3.92; p⬍.001; s ranging from ⫺.14 to ⫺.12),
and, most importantly, to the interaction term (CR ⫽⫺2.23; p⬍
.05; s ranging from ⫺.09 to ⫺.07). Based on these results as well
as the good fit indices of the models, we can conclude that the
findings on predictors of meaning in life in the total sample hold
true regardless of participants’ cultural background.
Discussion
The present study shows that the beneficial effect of generative
concern on meaning in life was partly mediated by generative
goals. Moreover, the effect of generative goals on meaning in life
was moderated by Machiavellianism. That is, for individuals with
low or medium Machiavellianism generative goals were more
strongly associated with meaning in life than for individuals with
high Machiavellianism. This pattern was found to be equivalent in
elderly participants from four cultures: Cameroon, China, the
Czech Republic, and Germany. These cultures were shown to
differ in value orientations: As expected, German participants
scored lowest for conservation, Cameroonian and Chinese partic-
ipants scored highest, and Czech participants ranged in-between.
For openness to change values, German participants scored highest
followed by Chinese participants who, in turn, scored higher than
Cameroonian and Czech participants. This relatively low openness
to change value orientation in the Czech sample was unexpected.
In sum, however, value orientations show that the cultural groups
selected represent a variety of psychologically defined cultural
orientations.
Generative Goals as Mediator Between Generative
Concern and Meaning in Life
The mediation assumption which predicted that the relationship
between generative concern and meaning in life is partly mediated
by generative goals was confirmed in all cultural samples. That is,
a positive attitude toward generativity contributes to meaning in
life by spawning generative goals. Goals are effective in providing
individuals with direction in life (e.g., Frazier et al., 2007) and,
particularly for the elderly, are seen by some authors to be an
integral part of well-being (Ryff, 1989).
This result also renders additional support to an essential pre-
diction of McAdams and de St. Aubin’s (1992) generativity model
in that it cross-culturally confirms the link between generative
concern and generative goals (Hofer et al., 2008;McAdams et al.,
1993). However, the mediation model as delineated above was
qualified by Machiavellianism which was used as a proxy variable
for belief in the species: The cynical, selfish, and manipulative
view on others that characterizes Machiavellianism opposes the
fundamentally positive view of mankind as evolving to the better
(see Busch & Hofer, 2012). As hypothesized, belief in the species
was found to moderate the association between generative goals
and meaning in life. That is, individuals low in belief in the species
do not benefit from generative goal pursuit in terms of meaning in
Figure 3. The mediation of generative goals between generative concern
and meaning in life is moderated by Machiavellianism. Note: Coefficients
are given as follows: (b/SE).
Figure 4. Effects of generative goals on meaning in life moderated by
level of Machiavellianism.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
772 HOFER ET AL.
life. Presumably, those high in Machiavellianism expect others not
to appreciate but to take advantage of their generative efforts and
hence find only little meaning in pursuing generative goals. This
interpretation is in line with Erikson’s (1963) argument that true
generativity with all its beneficial effects can only be achieved
when belief in the species is given. In line with this argument, then,
the decisive effect of belief in the species might not be that it
decreases generative goal setting but that it reduces the beneficial
effects that generative goal pursuit usually entails. Thus, the pres-
ent findings ought to draw attention to the neglected construct of
belief in the species and its position and function in the genera-
tivity model which has proved so helpful in systemizing research
on this central developmental task in adulthood.
Limitations and Outlook
In sum, analyses strongly confirmed our main hypotheses by
showing that, independent of culture, generative goals mediate the
relationship between concern for generativity and meaning in life
but only if the tendency to view people as objects to be manipu-
lated and exploited is low. Thus, although the study sheds light on
the relationship between facets of generativity and well-being in
old age, some limitations of the study have to be acknowledged.
In our study, altruistic goals were applied as a measurement for
generative strivings. Even if a considerable overlap between altru-
istic and generative goals can easily be assumed, a more direct
measure of generative goals might be used in future studies. For
example, an idiographic technique such as the personal striving
approach (Emmons & McAdams, 1991) may lead to a more
detailed assessment of generative goals (e.g., Hofer et al., 2008).
Again from a methodological point of view, a specific assessment
instrument for belief in the species has not yet been developed. Thus,
our operationalization of belief in the species with Machiavellianism
as its proxy, that is, lack of the conviction that generativity is worth-
while, might not have covered all facets of belief in the species: For
example, Cheng (2009) asked his participants about the respect they
perceive younger people to have for their generative efforts. However,
we are confident that high Machiavellian attitudes and low belief in
the species share some common ground (Busch & Hofer, 2012).
Nevertheless, future research might try to develop a specific instru-
ment for belief in the species or identify other proxy variables such as
social dominance orientation, that is, individual differences in how
much intergroup inequality is favored (e.g., Sidanius & Pratto, 2001).
Even if analyses indicated that participants’ cognitive perfor-
mance was associated with meaning in life across cultures, a note
has to be made on the meaningfulness of the test scores of the NAI.
The mean time of test trials in both speed tests (ZVT-G and LT-G)
was considerably higher among Cameroonian and Chinese than for
Czech and German participants. Probably, familiarity with test
material and procedures has affected test scores. We assume that
Cameroonian and Chinese participants have been less familiar
with speed tests and therefore scored lower in cognitive function-
ing. Thus, test scores are not suitable for cross-cultural mean
comparisons but can be used to examine relationships among
measurements (van de Vijver & Leung, 1997).
Some interesting findings concerning generativity and sociode-
mographic features occurred: First, age was positively correlated
with both generative concern and Machiavellianism. That is,
among the elderly the willingness to contribute to the development
of their junior ones seems to increase which underscores the
importance of generativity in old age. However, at the same time
attitudes which counteract this willingness seem to increase as
well. Whether this is due to a perceived decrease in abilities to act
generatively or to an increase in sobering experiences of unappre-
ciative reactions to generative efforts is a question that future
research might address.
Second, number of children was positively related to generative
concern and generative goals. Having children might thus be a
perpetual instigation for generative concern and generative goals
across the life span.
Next, future cross-cultural studies on generativity ought to in-
tegrate measurements on generative behavior to close the gap
between generative goals and perceived well-being. Above all,
goals that are successfully realized in behavioral generative strat-
egies are supposed to contribute to well-being. As cultural contexts
may differ from each other in given opportunities and constraints
to act generatively, an examination of cultural equivalence of
behavior seems to be indispensable. Preferably, future research on
generativity in various cultures will employ a longitudinal design
examining the dynamic nature of various facets of generativity and
successful aging.
To conclude, the study presents evidence for a significant rela-
tionship between generativity and well-being in old age across
highly diverse cultural groups. It was shown that the effect of
generative concern on meaning in life was partly mediated by
generative goals. Furthermore, the role of belief in the species, a
topic that has rarely been studied in the past, was examined:
Generative goals contributed to meaning in life only when not
opposed by low belief in the species, suggesting that lower levels
of belief in the species might prevent people from reaping the
beneficial effects of generative goal pursuit. In general, the study
exemplifies the benefits of a multicultural approach, permitting
meaningful insights about culture-specific developmental pro-
cesses but also about the extent to which developmental processes
can be generalized across cultures.
References
Ackerman, S., Zuroff, D. C., & Moskowitz, D. S. (2000). Generativity in
midlife and young adults: Links to agency, communion, and subjective
well-being. The International Journal of Aging and Human Develop-
ment, 50, 17–41. doi:10.2190/9F51-LR6T-JHRJ-2QW6
Arnett, J. (2008). The neglected 95%: Why American psychology needs to
become less American. American Psychologist, 63, 602–614. doi:
10.1037/0003-066X.63.7.602
Bandalos, D. L. (2002). The effects of item parceling on goodness-of-fit
and parameter estimate bias in structural equation modeling. Structural
Equation Modeling, 9, 78–102. doi:10.1207/S15328007SEM0901_5
Bereczkei, T., Birkas, B., & Kerekes, Z. (2010). The presence of others,
prosocial traits, Machiavellianism: A personality x situation approach.
Social Psychology, 41, 238–245. doi:10.1027/1864-9335/a000032
Brandtstädter, J., Rothermund, K., Kranz, D., & Kühn, W. (2010). Final
decentrations: Personal goals, rationality perspectives, and the aware-
ness of life’s finitude. European Psychologist, 15, 152–163. doi:
10.1027/1016-9040/a000019
Bühler, C., & Massarik, F. (1968). The course of human life. New York,
NY: Springer.
Busch, H., & Hofer, J. (2011). Identity, prosocial behavior, and generative
concern in German and Cameroonian Nso adolescents. Journal of Ad-
olescence, 34, 629–638. doi:10.1016/j.adolescence.2010.09.009
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
773
GENERATIVITY IN OLD AGE
Busch, H., & Hofer, J. (2012). Self-regulation and milestones of adult
development: Intimacy and generativity. Developmental Psychology, 48,
282–293. doi:10.1037/a0025521
Carstensen, L. L., Isaacowitz, D. M., & Charles, S. T. (1999). Taking time
seriously: A theory of socioemotional selectivity. American Psycholo-
gist, 54, 165–181. doi:10.1037/0003-066X.54.3.165
Chen, F. F. (2007). Sensitivity of goodness of fit indexes to lack of
measurement invariance. Structural Equation Modeling, 14, 464–504.
doi:10.1080/10705510701301834
Cheng, T.-S. (2009). Generativity in later life: Perceived respect from
younger generations as a determinant of goal disengagement and psy-
chological well-being. Journal of Gerontology: Psychological Sciences,
64B, 45–54. doi:10.1093/geronb/gbn027
Cheung, G. W., & Rensvold, R. B. (2002). Evaluating goodness-of-fit
indexes for testing measurement invariance. Structural Equation Mod-
eling, 9, 233–255. doi:10.1207/S15328007SEM0902_5
Christie, R., & Geis, F. (1970). Studies in Machiavellianism. New York,
NY: Academic Press.
Deci, E. L., & Ryan, R. M. (2000). The “what” and “why” of goal pursuits:
Human needs and the self-determination of behavior. Psychological
Inquiry, 11, 227–268. doi:10.1207/S15327965PLI1104_01
Diener, E. (1984). Subjective well-being. Psychological Bulletin, 95, 542–
575. doi:10.1037/0033-2909.95.3.542
Edwards, J. R., & Lambert, L. S. (2007). Methods for integrating moder-
ation and mediation: A general analytical framework using moderated
path analysis. Psychological Methods, 12, 1–22. doi:10.1037/1082-
989X.12.1.1
Emmons, R. A. (1986). Personal strivings: An approach to personality and
subjective well-being. Journal of Personality and Social Psychology, 51,
1058–1068. doi:10.1037/0022-3514.51.5.1058
Emmons, R. A. (2003). Personal goals, life meaning, and virtue: Well-
springs of a positive life. In C. L. M. Keyes & J. Haidt (Eds.), Flour-
ishing: Positive psychology and the life well-lived (pp. 105–128). Wash-
ington, DC: American Psychological Association. doi:10.1037/10594-
005
Emmons, R. A., & McAdams, D. P. (1991). Personal strivings and motive
dispositions: Exploring the links. Personality and Social Psychology
Bulletin, 17, 648–654. doi:10.1177/0146167291176007
Erikson, E. H. (1963). Childhood and society (2nd ed.). New York, NY:
Norton.
Erikson, E. H., Erikson, J. M., & Kivnick, H. Q. (1986). Vital involvement
in old age. New York, NY: Norton.
Fisher, B. J. (1995). Successful aging, life satisfaction, and generativity in
later life. The International Journal of Aging and Human Development,
41, 239–250. doi:10.2190/HA9X-H48D-9GYB-85XW
Frazier, L. D., Newman, F. L., & Jaccard, J. (2007). Psychosocial outcomes
in later life: A multivariate model. Psychology and Aging, 22, 676–689.
doi:10.1037/0882-7974.22.4.676
Fung, H. H., Carstensen, L. L., & Lutz, M. A. (1999). Influence of time on
social preferences: Implications for life-span development. Psychology
and Aging, 14, 595–604. doi:10.1037/0882-7974.14.4.595
Grossbaum, M. F., & Bates, G. W. (2002). Correlates of psychological
well-being at midlife: The role of generativity, agency and communion,
and narrative themes. International Journal of Behavioral Development,
26, 120–127. doi:10.1080/01650250042000654
Hofer, J., Busch, H., Bond, M. H., Campos, D., Li, M., & Law, R. (2010).
The implicit power motive and sociosexuality in men and women:
Pancultural effects of responsibility. Journal of Personality and Social
Psychology, 99, 380–394. doi:10.1037/a0020053
Hofer, J., Busch, H., Bond, M. H., Kärtner, K., Kiessling, F., & Law, R.
(2010). Is self-determined functioning a universal prerequisite for
motive-goal congruence? Examining the domain of achievement in three
cultures. Journal of Personality, 78, 747–780. doi:10.1111/j.1467-6494
.2010.00632.x
Hofer, J., Busch, H., Bond, M. H., Li, M., & Law, R. (2010). Effects of
motive-goal congruence on well-being in the power domain: Consider-
ing goals and values in a German and two Chinese samples. Journal of
Research in Personality, 44, 610–620. doi:10.1016/j.jrp.2010.08.001
Hofer, J., Busch, H., Chasiotis, A., Kärtner, J., & Campos, D. (2008).
Concern for generativity and its relation to implicit power motivation,
generative goals, and satisfaction with life: A cross-cultural investiga-
tion. Journal of Personality, 76, 1–30. doi:10.1111/j.1467-6494.2007
.00478.x
James, J. B., & Zarrett, N. (2005). Ego integrity in the lives of older
women: A follow-up of mothers from the Sears, Maccoby, and Levin
(1951). Patterns of child rearing study. Journal of Adult Development,
12, 155–167. doi:10.1007/s10804-005-7084-y
Jones, D. N., & Paulhus, D. L. (2009). Machiavellianism. In M. R. Leary
& R. H. Hoyle (Eds.), Handbook of individual differences in social
behavior (pp. 93–108). New York, NY: Guilford Press.
Jowell, R., & the Central Coordinating Team. (2007). European social
survey 2006/2007 (technical reports). London, UK: Centre for Compar-
ative Social Surveys, City University.
Keller, H. (2007). Cultures of infancy. Mahwah, NJ: Erlbaum.
Kline, R. B. (1998). Principles and practice of structural equation mod-
eling. New York, NY: Guilford Press.
Kotre, J. (1996). Outliving the self: How we live on in future generations.
New York, NY: Norton.
Kruse, A., & Wahl, H.-W. (2009). Zukunft Altern: Individuelle und gesell-
schaftliche Weichenstellungen [The future of aging: Individual and
societal directions]. Heidelberg, Germany: Spektrum.
Lang, F. R., & Carstensen, L. L. (2002). Time counts: Future time per-
spective, goals, and social relationships. Psychology and Aging, 17,
125–139. doi:10.1037/0882-7974.17.1.125
Lawton, M. P., & Brody, E. M. (1969). Assessment of older people:
Self-maintaining and instrumental activities of daily living. The Geron-
tologist, 9, 179–186. doi:10.1093/geront/9.3_Part_1.179
Little, T. D., Cunningham, W. A., Shahar, G., & Widaman, K. F. (2002).
To parcel or not to parcel: Exploring the question, weighing the merits.
Structural Equation Modeling, 9, 151–173. doi:10.1207/
S15328007SEM0902_1
MacDermid, S. M., Franz, C. E., & De Reus, L. A. (1998). Generativity:
At the crossroads of social roles and personality. In D. P. McAdams &
E. de St. Aubin (Eds.), Generativity and adult development: How and
why we care for the next generation (pp. 181–226). Washington, DC:
American Psychological Association. doi:10.1037/10288-006
Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications
for cognition, emotion, and motivation. Psychological Review, 98, 224–
253. doi:10.1037/0033-295X.98.2.224
Mbaku, J. M. (2005). Culture and customs of Cameroon. Westport, CT:
Greenwood Press.
McAdams, D. P. (2000). Attachment, intimacy, and generativity. Psycho-
logical Inquiry, 11, 117–120. doi:10.1207/S15327965PLI1102_02
McAdams, D. P., & de St. Aubin, E. (1992). A theory of generativity and
its assessment through self-report, behavioral acts, and narrative themes
in autobiography. Journal of Personality and Social Psychology, 62,
1003–1015. doi:10.1037/0022-3514.62.6.1003
McAdams, D. P., de St. Aubin, E., & Logan, R. L. (1993). Generativity
among young, midlife, and older adults. Psychology and Aging, 8,
221–230. doi:10.1037/0882-7974.8.2.221
McAdams, D. P., Diamond, A., de St. Aubin, E., & Mansfield, E. (1997).
Stories of commitment: The psychosocial construction of generative
lives. Journal of Personality and Social Psychology, 72, 678–694.
doi:10.1037/0022-3514.72.3.678
McAdams, D. P., Hart, H. M., & Maruna, A. S. (1998). The anatomy of
generativity. In D. P. McAdams & E. de St. Aubin (Eds.), Generativity
and adult development: How and why we care for the next generation
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
774 HOFER ET AL.
(pp. 7–43). Washington, DC: American Psychological Association. doi:
10.1037/10288-001
McHoskey, J. W. (1999). Machiavellianism, intrinsic versus extrinsic
goals, and social interest: A self-determination theory analysis. Motiva-
tion and Emotion, 23, 267–283. doi:10.1023/A:1021338809469
Nunnally, J. C. (1978). Psychometric theory. New York, NY: McGraw-
Hill.
Oswald, W. D., & Fleischmann, U. M. (1997). Das Nürnberger-Alters-
Inventar NAI [Nuremberg Aging Inventory NAI]. Göttingen, Germany:
Hogrefe.
Pepitone, A. (1976). Toward a normative and comparative biocultural
social psychology. Journal of Personality and Social Psychology, 34,
641–653. doi:10.1037/0022-3514.34.4.641
Peterson, B. E. (2006). Generativity and successful parenting: An analysis
of young adult outcomes. Journal of Personality, 74, 847–869. doi:
10.1111/j.1467-6494.2006.00394.x
Peterson, B. E., Smirles, K. A., & Wentworth, P. A. (1997). Generativity
and authoritarianism: Implications for personality, political involvement,
and parenting. Journal of Personality and Social Psychology, 72, 1202–
1216. doi:10.1037/0022-3514.72.5.1202
Pöhlmann, K., & Brunstein, J. C. (1997). Goals: Ein Fragebogen zur
Erfassung von Lebenszielen [Goals: A questionnaire assessing life
goals]. Diagnostica, 43, 63–79.
Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for
estimating indirect effects in simple mediation models. Behavior Re-
search Methods, Instruments, & Computers, 36, 717–731. doi:10.3758/
BF03206553
Preacher, K. J., Rucker, D. D., & Hayes, A. F. (2007). Addressing mod-
erated mediation hypotheses: Theory, methods, and prescriptions. Mul-
tivariate Behavioral Research, 42, 185–227. doi:10.1080/
00273170701341316
Rowe, J. W., & Kahn, R. L. (1998). Successful aging. New York, NY:
Pantheon.
Ryff, C. D. (1989). Beyond Ponce de Leon and life satisfaction: New
directions in quest of successful ageing. International Journal of Behav-
ioral Development, 12, 35–55. doi:10.1177/016502548901200102
Schoklitsch, A., & Baumann, U. (2011). Measuring generativity in older
adults: The development of new scales. GeroPsych: The Journal of
Gerontopsychology and Geriatric Psychiaty, 24, 31–43. doi:10.1024/
1662-9647/a000030
Schwartz, S. H. (1992). Universals in the content and structure of values:
Theoretical advances and empirical tests in 20 countries. In M. Zanna
(Ed.), Advances in experimental social psychology (Vol. 25, pp. 1–65).
Orlando, FL: Academic Press.
Sheldon, K. M., & Kasser, T. (2001). Getting older, getting better? Per-
sonal strivings and psychological maturity across the life span. Devel-
opmental Psychology, 37, 491–501. doi:10.1037/0012-1649.37.4.491
Sidanius, J., & Pratto, F. (2001). Social dominance: An intergroup theory
of hierarchy and oppression. Cambridge, UK: Cambridge University
Press.
Staudinger, U. M., & Fleeson, W. (1996). Self and personality in very old
age: A sample case of resilience? Development and Psychopathology, 8,
867–885. doi:10.1017/S0954579400007471
Steger, M. F., Frazier, P., Oishi, S., & Kaler, M. (2006). The meaning in
life questionnaire: Assessing the presence and search for meaning in life.
Journal of Counseling Psychology, 53, 80–93. doi:10.1037/0022-0167
.53.1.80
Tornstam, L. (1997). Gerotranscendence: The contemplative dimension of
aging. Journal of Aging Studies, 11, 143–154. doi:10.1016/S0890-
4065(97)90018-9
Triandis, H. C. (1996). The psychological measurement of cultural syn-
dromes. American Psychologist, 51, 407–415. doi:10.1037/0003-066X
.51.4.407
Vaillant, G. E., & Milofsky, E. (1980). Natural history of male psycho-
logical health: IX. Empirical evidence for Erikson’s model of the life
cycle. The American Journal of Psychiatry, 137, 1348–1359.
van de Vijver, F. J. R., & Leung, K. (1997). Methods and data analysis for
cross-cultural research. Newbury Park, CA: Sage.
Van De Water, D. A., & McAdams, D. P. (1989). Generativity and
Erikson’s “belief in the species.” Journal of Research in Personality, 23,
435–449. doi:10.1016/0092-6566(89)90013-5
Waddell, E. L., & Jacobs-Lawson, J. M. (2010). Predicting positive well-
being in older men and women. The International Journal of Aging and
Human Development, 70, 181–197. doi:10.2190/AG.70.3.a
Received August 15, 2013
Revision received July 3, 2014
Accepted July 7, 2014 䡲
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
775
GENERATIVITY IN OLD AGE
