Developmental Regulation Across Adulthood: Primary and Secondary Control of Age-Related Challenges

Abstract

This study addresses developmental regulation in adults at different ages. A conceptual model of optimization in primary and secondary control across the life span (OPS model) is presented, and predictions about age differences in developmental regulation are derived. Developmental goals, expectations about goal attainment and control, control strategies, life satisfaction, and age identification were assessed in a sample of 510 young, middle-aged, and old adults. At increasing age, the participants expressed (a) greater awareness of a reduced potential for growth and control. (b) increased focus on age-appropriate goals for primary control striving, (c) more goals directed at the avoidance of developmental losses and fewer goals aimed at developmental gains, and (d) a stronger tendency for compensatory secondary control, as shown in greater goal flexibility, more satisfaction with present life, and identification with younger age groups.

Full text

Developmental Psychology
1997,
Vol. 33, No. 1, 176-187
Copyright 1997 by the American Psychological Association, Inc.
0012-l649/97/$3.00
Developmental Regulation Across Adulthood: Primary and
Secondary Control of Age-Related Challenges
Jutta Heckhausen
Max Planck Institute for Human Development and Education
This study addresses developmental regulation in adults at different ages. A conceptual model of
optimization in primary and secondary control across the life span (OPS model) is presented, and
predictions about age differences in developmental regulation are derived. Developmental goals,
expectations about goal attainment and control, control strategies, life satisfaction, and age identifica-
tion were assessed in a sample of 510 young, middle-aged, and old adults. At increasing age, the
participants expressed (a) greater awareness of a reduced potential for growth and control, (b)
increased focus on age-appropriate goals for primary control striving, (c) more goals directed at the
avoidance of developmental losses and fewer goals aimed at developmental gains, and (d) a stronger
tendency for compensatory secondary control, as shown in greater goal flexibility, more satisfaction
with present life, and identification with younger age groups.
This article addresses adults' developmental regulation in dif-
ferent periods of the adult life span. On the basis of a life
span theory of control (J. Heckhausen & Schulz, 1995) and its
integration with a model of developmental regulation across the
life span (OPS model; J. Heckhausen & Schulz, 1993; Schulz &
Heckhausen, 1996), a set of predictions about age differences
in striving for developmental goals is developed and put to the
empirical test.
Model of Developmental Regulation
Developmental regulation comprises individuals' efforts to
influence their own actual development on the one hand and
their attempts to adapt psychologically to a given developmental
ecology as a constraining context to development on the other.
These two components of developmental regulation resemble
the more general concepts of primary and secondary control as
Portions of this work were presented at the 38th Congress of the
Deutsche Gesellschaft fur Psychologic, Trier, Germany, 1992, and at the
100th Annual Convention of the American Psychological Association,
Washington, DC, August 1992, in a paper coauthored with Richard
Schulz, to whom I am indebted for many years of debate and collabora-
tion on this topic. This article was completed while I was a fellow at
the Center for Advanced Study in the Behavioral Sciences, Stanford,
California. 1 am grateful for financial support provided by John D. and
Catherine T. MacArthur Foundation Grant 8900078.
I thank Anette Baumeister, Birgit Grabow, Jutta Hundertmark, and
Carsten Wrosch for their help with data collection and analyses and Paul
B.
Baltes and the life span development research group at the Max.
Planck Institute for Human Development and Education, Berlin, for
many collegial discussions. I am especially grateful to Laura L. Cars-
tensen, William W. Fleeson, and Richard Schulz for their thoughtful
comments on a draft of this article.
Correspondence concerning this article should be addressed to Jutta
Heckhausen, Max Planck Institute for Human Development and Educa-
tion, Lentzeallee 94, 14195 Berlin, Germany. Electronic mail may be
sent via Internet to heckhausen@mpib-berlin.mpg.de.
proposed by Rothbaum, Weisz, and Snyder (1982). Primary
control pertains to attempts to change the external world so that
it fits the needs and desires of the individual. Secondary control,
by contrast, targets the internal world of the individual in efforts
to "fit in with the world" (Rothbaum et al., 1982).
J. Heckhausen and Schulz
(1993,
1995; Schulz, Heck-
hausen, & Locher, 1991) have elaborated the conceptual scheme
of primary and secondary control and developed a life span
theory of control. According to this theory, primary control holds
functional primacy over secondary control in that secondary
control serves to protect the motivational resources for primary
control (see also Heckhausen & Schulz, 1994). Primary and
secondary control both serve two main requirements of human
behavior: the need to be selective and the need to cope with
failure. Human behavior needs to be selective because of the
enormous behavioral options available to the human species.
The need to cope with failure arises in principle from the same
basis.
A lack of predeterminedness in behavior brings about
many opportunities to fail. In fact, acquisitional processes often
rely on learning from failure. Thus, humans have to be able to
cope with the frustration and potential threats to self-esteem
and motivational resources that accompany regular experiences
of failure. Negative effects of failure on self-esteem and motiva-
tional resources are particularly consequential because they un-
dermine the individual's long-term capacities for effective
functioning.
Ontogenetic change throughout the human life span amplifies
the role of selectivity because potential developmental trajector-
ies imply even greater heterogeneity, and because based on accu-
mulated selective investment, even more is at stake. Moreover,
the need for failure compensation is also enhanced with age
because with increasing age more losses are encountered. The
notion of enhanced utility of selection and compensation pro-
cesses with increasing age is the cornerstone of life span re-
search, as conceptualized in Baltes and Baltes' model of selec-
tive optimization with compensation (P. B. Baltes & Baltes,
1990).
176

DEVELOPMENTAL REGULATION
177
J. Heckhausen and Schulz (1993) proposed a two-dimen-
sional model of optimization in primary and secondary control
(OPS model). The OPS model extends and elaborates Baltes
and Baltes's model of selective optimization with compensation
(P.
B. Baltes & Baltes, 1990; P. B. Baltes, Lindenberger, &
Staudinger, in press; Marsiske, Lang, P. B. Baltes, & Baltes,
1995) by integrating it with the life span theory of primary and
secondary control (J. Heckhausen & Schulz, 1995). The OPS
model conceptualizes developmental regulation as being com-
posed of the dimensions of primary and secondary control on
the one hand and selection and compensation on the other (J.
Heckhausen & Schulz, 1993). Four types of control strategies
are identified in this two-dimensional model.
First, selective primary control refers to focused investment
of resources such as time, effort, abilities, and skills in the
attainment of goals. In other words, selective primary control
is action that directly aims at attaining goals. Second, compensa-
tory primary control is required when the given internal re-
sources of the individual prove insufficient to attain a chosen
goal, as in young children or frail, elderly people. Typically,
compensatory primary control involves the recruitment of extra-
neous resources, such as others' help or technical aids, to supple-
ment own capacities (e.g., Marsiske et al., 1995). In addition, it
also includes the employment of new, unusual, and thus activity-
external action means to substitute lost or otherwise unavailable
skills inherent to the activity, such as in lipreading of hearing-
impaired persons (e.g., Thompson, 1995). Third, selective sec-
ondary control is directed at the internal world of the individual
to promote the volitional commitment to a chosen goal. Typical
examples are boosting the value of a chosen goal or heightened
perceptions of personal control (e.g., Gollwitzer & Kinney,
1989).
Fourth, compensatory secondary control is aimed at min-
imizing negative effects of failure on the individual's self-esteem
and motivational resources, such as perceived personal control
and self-efficacy (Bandura, 1982). This can be done by such
strategies as goal disengagement (e.g., Brandtstadter & Rother-
mund, 1994; Brandtstadter, Wentura, & Greve, 1993; Klinger,
1975) or strategic social comparison (e.g., Taylor & Lobel,
1989;
Wills, 1981; Wood, 1989). Finally, the effective use of
each of the four strategies is regulated by a higher order process:
optimization. Optimization balances the employment of the four
strategies according to the developmental opportunities and con-
straints so as to expand, maintain, and protect the long-term
potential for primary control.
Age-Related Challenges Across Adulthood
Age-related challenges to developmental regulation are mani-
fold and can be identified in various domains and at multiple
age transitions during adulthood. The present study addresses
challenges associated with age-graded developmental tasks
(Havighurst, 1952) and focuses on those challenges associated
with aging. Aging-related challenges can be seen as particularly
salient examples of the demands encountered by adults undergo-
ing developmental transitions at various ages during adulthood.
The changes associated with advanced age confront the individ-
ual with increasing constraints and losses in the biological realm
as well as with regard to social roles. Although there is some
scope for interindividual variability (e.g., Nelson & Dannefer,
1992) and intraindividual plasticity (e.g., Kliegl, Smith, &
Baltes, 1990), aging-related losses eventually are inescapable
for everyone. Consider, for instance, age-related constrictions
in social roles (e.g., mothering) associated with menopause.
The individual can hardly avoid such role changes. However,
she can still use secondary compensation by disengaging from
caregiving goals and substituting other generativity goals and
activities (e.g., teaching students) that so far have remained
unaffected by aging-related loss or that indeed represent unique
developmental tasks (Havighurst, 1952) of old age (e.g., grand-
parenting). Similarly, in the realm of work, most occupational
careers set fairly strict age limits for attaining further promotion
("developmental deadlines"; see J. Heckhausen, in press; J.
Heckhausen & Fleeson, 1993). The individual has to take these
into account when striving for advancement and disengaging
from previous goals. In a similar way, biological changes related
to aging probably prompt most individuals to start being con-
cerned about impending decline in health (e.g., Harris & Associ-
ates,
1981). Some individuals consequently may attempt to pre-
vent or at least slow health decline by special compensatory
primary control (e.g., diets, exercise). However, biological
aging will eventually take its inevitable toll and thus has to be
accepted as a fact of life.
The basic assumption is that individuals in general adapt their
regulatory strategies to the opportunities and constraints en-
countered in a given situation. In this way, older adults do not
qualitatively differ from younger adults but respond to a more
constrained developmental ecology, which is characterized by
developmental losses more than by gains. The individual will
strive to maintain and extend primary control, if at all possible
(J. Heckhausen & Schulz, 1993, 1995). Only when opportuni-
ties for active influence on development are severely constrained
will the individual disengage from respective primary control
strivings and reorient control striving in line with available op-
portunities for primary control. Hence, challenges related to
aging are more likely to elicit compensatory-type control strate-
gies,
in particular compensatory secondary control. However,
primary control will also be pursued and should be expressed
in two ways. First, primary control in old age should selectively
address age-appropriate developmental tasks, such as health care
and public or community engagement (Nurmi, 1992). Second,
selective primary control should focus on the avoidance of de-
velopmental losses rather than on the striving for developmental
gains.
Developmental Goals as Organizers
of Developmental Regulation
The expectations outlined previously about general control
strategies preferred by adults confronted with aging are investi-
gated with regard to key indicators of developmental regulation,
namely characteristics of behavior directed at developmental
goals.
The concept of developmental goals is based on the notion
that individuals act as producers of their own development (Ler-
ner & Busch-Rossnagel, 1981). Although this notion is widely
accepted in developmental psychology (see, Brandtstadter, in
press),
little empirical work has addressed active attempts to
shape one's own development.
Goal-related phenomena have been investigated both out-

178
HECKHAUSEN
side of the developmental field (Cantor & Fleeson, 1994;
Emmons, 1986; Emmons & Diener, 1986; Gollwitzer, 1987;
Klinger, 1975; Klinger, Barta, & Maxeiner, 1980; Little, 1983;
Palys & Little, 1983; Wadsworth & Ford, 1983) and within
developmental research (see review in BrandtstSdter, in press;
Cantor & Fleeson, 1991; Cantor, Norem, Niedenthal, Langs-
ton, & Brower, 1987; Carstensen, 1993; Nurmi, 1992). Pre- or
semi structured modes of assessing goals were mostly used in
developmental research (e.g., BrandtstSdter & Renner, 1990;
Cantor et al., 1987). Nurmi (1992), however, has investigated
life goals and concerns in adults by using an idiographic open-
ended approach. His adult sample at different age levels nomi-
nated life goals that reflected the respective age-appropriate de-
velopmental tasks (Havighurst, 1952). ^
The present approach views developmental goals as organiz-
ers of developmental regulation. Developmental goals provide
an important medium through which developmental regula-
tion—that is, optimizing selection and compensation via pri-
mary and secondary control—has an impact on behavior. Devel-
opmental goals are future outcomes of development that the
individual tries to attain within a medium time range (i.e., 5-
10 years). Developmental goals can either reflect the striving
for gains or the avoidance of losses. An idiographic (open-
ended) assessment was chosen to capture the specificity, con-
creteness, and ecological validity of an individual's develop-
mental goals (see also Nurmi, 1992).
The individual will select developmental goals appropriate to
a given developmental or life-course challenge. This often im-
plies disengagement from previous goals that have become unat-
tainable and pursuit of new goals, for which action opportunities
are becoming available in an upcoming developmental ecology.
Thus,
compensatory secondary and selective primary control
work hand in hand.
When the developmental challenge is characterized by a low
potential for personal control, enhanced striving for primary
control becomes dysfunctional. As a first step, this situation will
usually prompt a shift in focus from striving for gains to the
prevention of losses. However, one can eventually expect that
individuals will resort to compensatory secondary control. Com-
pensatory secondary control strategies need to address two neg-
ative consequences of losses in primary control. First, negative
affect results from the frustration of primary control goal striv-
ing. Such frustration can be avoided by disengaging from the
respective goal. Indicators of such disengagement is the tendency
to adjust flexibly one's goals (Brandtstadter & Renner, 1990)
and the satisfaction with one's present life situation. Second,
feelings of self-esteem and general perceptions of self-worth
are threatened by losses in primary control. Such self-related
threats can be counteracted by a reinterpretation of one's own
status relative to others, by self-serving attributional patterns,
and by other strategies. In this research, two indicators of subjec-
tive age perception were chosen because of their relevance to
aging-related losses: positively biased perceptions of one's own
status with regard to age (subjective age identification) and a
revision of desired outcomes in view of one's own standing
(perceived prime of life).
Predictions
The predictions raised here are addressing age-related change
throughout adulthood. This implies predicted differences be-
tween young and middle-aged adults and between middle-aged
and old adults. For reasons of parsimony and salience, the pre-
dictions are phrased as being about differences between older
and younger adults.
Developmental regulation in older compared with younger
adults should have the following characteristics. First, it is ex-
pected that older compared with younger adults will express an
awareness of the decreased potential for developmental growth.
Such an awareness is the subjective precondition for activating
control strategies to meet the aging-related developmental chal-
lenge. An awareness of decreased potential for developmental
growth should be indicated by less optimistic expectations about
the likelihood of attaining goals and a decreased perception of
personal control over goal attainment.
Second, selective primary control in older adults should be
reflected in maintained primary control striving, which selec-
tively focuses on age-appropriate developmental goals. This goal
selection should be accompanied by a disengagement from pre-
vious,
now less attainable (or unattainable) goals. Moreover,
older adults' selection of developmental goals should imply a
shift in the ratio between goals pertaining to the striving for
developmental gains and goals representing an avoidance of
losses. Thus, it was expected that older compared with younger
adults would nominate a greater number of goals representing
avoiding losses in comparison to goals reflecting striving for
gains.
Third, compensatory secondary control in older adults should
be expressed in four indicators: first, a greater flexibility in
disengaging from goals; second, a higher satisfaction with the
current life situation; third, a greater tendency to identify with
younger age groups, both in terms of one's subjective age status
and the normative age ascribed to one's own developmental
goals;
fourth, the perception that the prime of life is at relatively
higher ages.
Method
Participants
All participants (N = 510) were volunteers, recruited by newspaper
advertisements in Berlin. They were paid 30 DM (approximately $20
in the United States) for the 1.5 h session. Of the sample, 162 were
classified as young (age range = 20-35 years; M = 28.66, SD - 4.29),
173 as middle aged (age range - 40-55 years; M = 47.80, SD = 4.93),
and 175 as old (older than 60 years; M =
Q6.Z1,
SD = 5.32). Regarding
place of origin, 255 were from East Berlin and 255 were from West
Berlin. There were 255 women and 255 men. Regarding occupational
status,
125 were teachers, 131 were engineers, 128 were clerks, and 126
were skilled and semiskilled workers. This breakdown according to age,
east-west origin, sex, and socioeconomic status made up a subject
design involving 48 cells. The total sample of 510 participants was
evenly distributed across the 48 cells of the design, yielding an average
of 10.6 (SD = 1.0) participants per cell, with only minor deviations
from a completely balanced design.
Procedure
Data collection took place between April 1991 and February 1992.
Subjects were invited to group sessions at the Max Planck Institute
for Human Development and Education. Groups consisted of 10 to 15
respondents. Participants were given a booklet with the various question-

DEVELOPMENTAL REGULATION
179
naires, arranged
in
fixed order with open-ended questions
(on
develop-
mental goals) preceding item ratings. This article reports part
of a
larger
study involving
a
more extensive
set of
variables. Only those variables
are addressed here that are relevant
for
predictions raised
in
the
introduc-
tion. Additional questions addressed perceptions about controllability
and personal influence
on
various causal factors involved
in the
life
domains
of
family, work,
and
community
(die
Control, Agency, Means-
End
in
Adulthood Questionnaire, CAMAQ;
J.
Heckhausen
&
Hundert-
mark, 1995), attitudes toward German reunification,
and
self-related
and stereotypic conceptions about East
and
West German psychological
characteristics.
Participants were requested
to
give
the
five "most important personal
hopes,
plans,
and
goals
for the
next
5 to 10
years" (instruction similar
to Nurmi, 1992).
For
each
of the
nominated goals, subjects were also
asked about probability
of
attainment
("In
your view,
how
likely
is it
that this goal will
be
realized?" [0%-100%]); personal control
("How
much influence
do you
think
you
have over attaining this goal?"
[1 =
very little influence,
5 =
very much influence]);
and the
normative
age
of goal (subjects were asked
to
estimate
the
typical
age at
which most
other people would pursue
the
respective goal).
The
latter measure
of
normative
age of
goal
was
added
to the
design
as an
additional, more
specific goal-related indicator
of
subjective
age
identification
and was
assessed
by
mailed questionnaire
4
months after
the
main study.
A
total
of
457
participants returned these ratings. This sample
did not
differ
from
the
original sample
in
terms
of the
design variables.
The Tenacious Goal Pursuit
and
Flexible Goal Adjustment Scales
(TEN-FLEX; Brandtstadter
&
Renner,
1990)
were presented
as a mea-
sure
of
goal pursuit (selective primary control)
and
goal disengagement
(compensatory secondary control;
1 =
disagree very much,
5 -
agree
very much).
A
sample item from
the
Tenacious Goal Pursuit Scale
is,
"When
I
have
set my
mind
on
something,
I do not get
discouraged even
by great difficulties,''
and
for
the Flexible Goal Adjustment Scale,' 'Even
in great misfortune
I
often find meaning." Subjects were asked about
their present life satisfaction
('
'How satisfied
are you
with your life
at
present?";
1 = not at all
satisfied,
7 =
very much satisfied).
A
set of
questions about subjective
age
identification
('
'The Ages
of
Me";
Kastenbaum, Derbin, Sabatini,
& Am, 1981)
were asked, includ-
ing
a
scale made
up of
five aspects
of
subjective
age ("the age I
feel
like,"
"the age I
look like,"
"the age
resembling
my
interests
and
activities,"
"the age
unacquainted people would ascribe
to me," and
"the
age my
friends would ascribe
to me").
Subjects responded
to
these items
in
terms
of
rating these aspects
of
subjective
age as
younger,
same
age,
or
older than myself (1
=
much older than
myself,
3 =
same
age
as myself, 5 =
much younger than myself). Note that another
indicator
of
subjective
age
identification
was
gathered
in
terms
of
esti-
mates
of
normative
age of
personal developmental goals. Subjects were
asked whether the prime
of
life
was
in
the past
(age
estimate), present,
or
future
(age
estimate). Finally, some basic sociodemographic information
(e.g.,
age,
family status, current residence, number
and age of
children,
education, occupational training, current occupational position) were
gathered.
Results
The Results section
is
organized according
to the
scheme
of
predictions presented previously about developmental regulation
in older compared with younger adults. Because
of
the relatively
large sample,
a
criterion
of p < .01
(rather than
p < .05) was
chosen
as a
cutoff
for
significant findings.
The
between-subjects
factors
of
sex, occupational status,
and
East Berlin versus West
Berlin origin serve
as
a
means
for
investigating the generalizabil-
ity
of
age
differences across gender, socioeconomic,
and
socio-
historical settings.
A
more detailed discussion
of
differences
between East Berliners
and
West Berliners
is
provided
in J.
Heckhausen (1994,
in
press).
Awareness
of
Decreased Potential
for
Developmental
Growth
Probability
of
goal attainment.
An
analysis
of
variance
based
on all
between-subjects factors uncovered
a
significant
main effect
for age in the
probability
of
goal attainment,
F(2,
455)
=
7.54,
p -
.001,
TJ
2
=
.032.
Follow-up analyses showed
that,
as
expected,
the old (M
—
66.9%)
and the
middle-aged
(M
=
65.4%) adults were less optimistic about
the
probability
of goal attainment than
the
young adults
(M =
72.3%),
/(400.82)
=
-4.23,
p < .001.
Personal control of goal attainment.
An
ANOY\ involving
all between-subjects factors yielded
a
significant main effect
of
age,
F(2, 462) =
37.99,
p =
.001,
TJ
2
=
.141. Follow-up tests
indicated that,
in
accordance with
the
prediction,
the old
adults
(M
- 3.20)
perceived less personal control over attaining their
goals than
the
middle-aged adults
(M =
3.48), r(322.74)
=
-3.22,
p < .001, who, in
turn, ascribed
to
themselves less
control than
the
young adults
(M =
3.92), r(321.17)
=
-6.55,
p
< .001.
Selective Primary Control
Content domain of developmental goals. Goals were coded
according
to a
highly differentiated system
of
29
categories,
for
which interrater agreement, assessed
on 10% of the
goal
cod-
ings,
was
92%
(K =
.91).
For
analysis, various categories were
collapsed into
a
scheme
of
six
to
avoid
low
frequencies
in
some
of
the
goal categories.
The
classification
of
categories into
the
six summary categories
was
based
on
their association with
basic domains
of
life (e.g., work, family).
The six
summary
categories were
(a)
work-related goals
(job
characteristics,
edu-
cation, occupational training);
(b)
family-related goals (partner,
children, parents);
(c)
goals directed
at
civic
and
community
issues (economy, social welfare, politics, environment);
(d)
goals associated with personal health;
(e)
goals directed
at per-
sonal financial welfare;
and (f)
goals addressing free time
and
leisure.
A
3 (age) X 2
(east-west)
X 2 (sex) x 4
(occupation)
X
6 (goal categories) ANOVA
was
performed, with goal category
as
a
within-subjects factor. Because
the
Category
X Age
interac-
tion,
F(10, 2310) =
39.74,
p < .001, r\
2
= .140, and the
Category
x
East-West interaction,
F(5, 2310) = 7.27, p <
.001,
T]
2
=
.012, were significant, analyses were performed sepa-
rately
for
each
of
the
six
goal categories. Significant differences
between
age
groups
are
displayed
in
Figure
1
(Top;
decreasing
trajectories,
and
(Bottom; increasing trajectories).
The work-related goals yielded
a
significant main effect
of
age,
F(2, 462) =
159.07,
p <
.001,
r\
2
-
.408,
and
east-west
origin,
F(
1,462)
=
21.72,/?
<
.001,
rj
2
=
.045. Figure
1 (Top)
shows
the
mean frequencies
(and
their standard errors)
of
goal
nominations across
age for
those goals with significantly
de-
creasing
age
trajectories.
As can be
seen
in
Figure
1
(Top),
younger
and
middle-aged adults nominated more work-related
goals than older adults, t(504)
=
20.56,
p <
.001.
East Berliners
(M
= .87)
expressed more work-related goals than West Berli-

180 HECKHAUSEN
1,2
1 -
0,8-
= 0,4
1
e
Z
0,2-1
- ^
J.
-
\
*"-<> family
\ """Y
finances
\
2 work
young
middle
Age Group
old
1,2
1-
0,8-
0,6-
a
.2
'-5
3 0,4-
s
o
z
0,2-
...,^ health
leisure
community
1
young
1
middle
Age Group
old
Figure 1. (Tbp) Mean nomination frequencies for developmental goals
with increasing age trajectories. (Bottom) Mean nomination frequencies
for developmental goals with decreasing age trajectories.
ners (M = .64). Financial goals were more frequently endorsed
by young adults compared with middle-aged and old adults,
F(2,
462) - 20.34, p < .001, rj
2
= .081, /(270) = 5.77, p <
.01 (see Figure 1, Top). The Age x Occupation interaction was
also significant, F(6, 462) = 2.95, p < .01, rj
2
= .037, and
implied that the higher frequency of financial goals of the young
compared with the middle-aged and old adults pertained only
to the teachers and clerks in the sample. Family-related goals
were reported more frequently by younger compared with mid-
dle-aged and older adults, F{2, 462) = 23.29, p < .001, t]
2
=
.092,
/(508) = 6.67, p < .01 (see Figure 1, Top). Health-
related goals yielded a significant main effect for age, F(2,
462) = 46.57, p < .001, r/
2
= .168, and for east-west origin,
F(\, 462) = 7.00, p < .01, rj
2
= .015. Figure 1 (Bottom)
exhibits the mean frequency of goal nominations across age for
those goals that showed an increasing trajectory with age. As
can be seen in Figure 1 (Bottom), older adults expressed more
health-related goals than middle-aged adults, Z(346) = 4.22, p
< .001, and middle-aged adults more than young adults, /(333)
= 5.74, p < .001. East Berliners (M = .60) reported fewer
health-related goals than West Berliners (Af = .74). The fre-
quency of leisure-related goals differed significantly across age,
F(2,
462) = 13.09, p < .001, rj
2
= .054, and sex, F(l, 462)
= 12.12,/? < .001, j]
2
= .026. As shown in Figure 1 (Bottom),
older adults reported more leisure-related goals than middle-
aged and young adults, ?(508) = 4.83,p <
.001.
Men expressed
fewer leisure-related goals than women (men: M = .49, women:
M = .70). Community-related goals were less frequent among
the young compared with the middle-aged and old adults, F(2,
452) = 12.66,p < .001, -q
1
= .052, r(508) = -4.46, p < .001
(see Figure 1, Bottom).
Avoiding losses versus striving for gains. In addition to the
categorization with regard to life domains, the goals nominated
by the subjects were also coded with regard to whether they
reflected attempts to avoid losses (e.g., "not become unem-
ployed," "keep one's apartment," "stay healthy") or to strive
for gains (e.g., "get a better paid job," "increase one's knowl-
edge about other cultures," "intensify one's relationship to a
child").
A 3 (age) X 2 (east-west) x 2 (sex) x 4 (occupation) X 2
(gain-loss) ANOVA was conducted, with gain-loss as a within-
subjects factor. There was a significant main effect for gain-
loss,
F(l, 462) = 381.18,/? < .001, r)
2
= .452, indicating that
striving for gains was much more frequent than avoiding losses
(striving for gains: M = 3.28, avoiding losses: M =
1.47).
Moreover, the Gain-Loss
X
Age, F(2, 462) = 24.28, p < .001,
r)
2
= .095, and the Gain-Loss x Occupation, F(3, 462) =
3.38, p < .01,
T)
2
= .021, interactions were significant. Figure
2 displays the significant age differences in terms of age group
means and standard errors. Separate follow-ups for gain-striving
goals and loss-avoiding goals revealed that, as predicted, the
o
O
s
J;
young middle
Age Group
r
old
Figure
2.
Striving for gains and avoiding losses in developmental goals
across age groups.

DEVELOPMENTAL REGULATION
181
young adults mentioned significantly more gain-striving goals
than the middle-aged adults, f(232.08) = 5.54, p < .001, and
old adults reported significantly fewer gain-striving goals than
middle-aged and young adults, r(508) - -5.29, p < .001.
Conversely, the younger adults generated fewer loss-avoiding
goals than the middle-aged adults, t(333) = -4.97, p < .001,
and the old adults mentioned more loss-avoiding goals than the
two younger age groups, f(5O8) = 3.23, p < .001. The signifi-
cant Gain-Loss X Occupation interaction implied that the
adults in occupations involving higher education (i.e., teachers,
engineers) reported more instances of striving for gains than
those in positions requiring less education (i.e., workers,
clerks),
/(508) = 3.17, p < .001 (high-education occupations:
M
—
3.44, low-education occupations: M = 3.11).
Compensatory Secondary Control
Tenaciousness of goal pursuit and flexibility of goal adjust-
ment A 3 (age) x 2 (east-west) x 2 (sex) X 4 (occupation)
X 2 (TEN-FLEX) multivariate ANO\ft (MANOVA) was per-
formed, with TEN-FLEX as a within-subjects factor measured
on the basis of 15 items for TEN and 15 items for FLEX. Various
interactions involving TEN-FLEX and between-subjects factors
were obtained: TEN-FLEX X Age, F(58, 462) = 3.73, p =
.001,
rj
2
= .110; TEN-FLEX X East-West, F(29,462) - 2.12,
p = .001, T)
2
= .066; TEN-FLEX X Occupation, F(87, 462)
= 2.93, p = .001,
TJ
2
= .086; TEN-FLEX x Age x East-West
X Sex, F(58, 462) = 1.76, p = .001,
T?
2
= .055. Therefore,
separate analyses for tenaciousness and flexibility were
conducted.
The MANOVA on tenaciousness ratings yielded significant
main effects for age, F(28, 462) == 5.47, p = .001, r}
2
= .149,
east-west, F( 14,462) =
2.08,
p =
.001,
TJ
Z
= .065, and occupa-
tion, F(42, 462) = 2.64, p = .001,
TJ
2
= .079. Although age
differences were significant as an overall between-subjects fac-
tor, specific age contrasts did not yield significant results to
support a decreasing trend.
1
Tenaciousness was greater in East Berliners (M - 52.40)
compared with West Berliners (M = 51.46). The occupation-
based differences were due to the fact that those whose occupa-
tions required less education (workers, clerks) expressed less
tenaciousness than those whose occupations required more
(teachers, engineers), f(5O8) = -2.74, p < .01.
The MANOVA on ratings of flexibility resulted in a signifi-
cant main effect for age, F(28, 462) = 2.03, p = .001,
TJ
2
=
.063,
east-west, F( 14, 462) = 2.77, p = .001, rj
2
= .085, and
sex, F(14, 462) = 2.52, p = .001, rj
2
= .079. Moreover, the
Age x East-West X Sex interaction was significant, F(28,
462) = 1.99, p = .001, r/
2
= .062. A follow-up of the significant
age effect and Figure 3 indicate that, as predicted, the old adults
expressed more flexible goal adjustment than the middle-aged
adults, r(346) = 1.95, p < .01, and the middle-aged adults
indicated more flexibility than the young adults, ((333) = 2.80,
p <
.01.
Moreover, West Berliners (M = 53.74) provided higher
flexibility ratings than East Berliners (M = 52.97). The signifi-
cant sex effect implied that men expressed greater flexibility in
goal adjustment than women (women: M =
52.93,
men: M =
53.78).
Regarding the three-way Age X East-West x Sex interaction,
55-
54-
52
51-
50-
49-
48
T
J> flexibility
tenaciousness
young
middle
Age Group
old
Figure 3. Flexibility of goal adjustment and tenaciousness of goal
pursuit across age groups.
mean differences were investigated to determine whether the
pattern of overall age differences remains constant across East
and West Berliners and across men and women. Age-related
increases were found for East and West Berlin women and men,
except that for the West Berlin men an increase in flexibility
was found only between young and middle adulthood and not
between middle-aged and old men.
Life satisfaction. A 3 (age) X 2 (east-west) X 2 (sex) X
4 (occupation) ANOVA was performed on the ratings of present
life satisfaction. Significant main effects were obtained for age,
F(2,
462) = 9.64, p - .001, 7?
2
= .04, and east-west, F(l,
462) =
47.91,
p = .001, rj
2
= .094. As expected, old adults (M
= 3.77) were more satisfied with their present life than young
(M = 3.46) and middle-aged (M = 3.34) adults, *(508) =
3.88, p < .001. Moreover, East Berliners (M = 3.50) were less
satisfied with their present lives than West Berliners (M = 3.84).
Subjective age identification. The 3 (age) X 2 (east-west)
X 2 (sex) X 4 (occupation) ANOVA, performed on the mean
ratings for the subjective age identification scale, revealed a
significant main effect for age, F(2, 462) =
55.51,
p = .001,
T)
1
= .194. Compared with their own age, the older adults (M
= 1.95) identified more with relatively younger people than the
middle-aged adults {M = 2.19), r(346) = 3.64,/? < .001, and
similarly, the middle-aged adults identified with—-compared
with their own age—relatively younger age groups than did the
young participants (M = 2.68), f(333) = 6.68, p < .001.
Normative age ascribed to developmental goal. Note that
these analyses are based on a subsample of the 457 subjects
who returned the mailed questionnaire about normative age of
goals.
A 3 (age) X 2 (east-west) X 2 (sex) X 4 (occupation)
ANOV^ was performed on the ratings of normative age of goals.
This analysis yielded a significant main effect of age, F(2, 409)
1
Factor analytic investigation of this inconsistent result revealed that
the Tenaciousness Scale comprised two factors with discrete loading
patterns.

182
HECKHAUSEN
70
60-
C
2 50
30-
20
participants'
age
,,..-•1 best age
.*-."-•" goal age
young
middle
Age Group
old
Figure
4.
Participants' actual age, normative age ascribed to goal, and
age perceived to be the prime of life across age groups.
= 83.79, p - .001, ij
2
= .291. Figure 4 displays the normative
goal ages given by the young, middle-aged, and old adults along
with their actual ages and the age they perceived as the prime
of life (see results reported later). The young adults ascribed
younger ages to their developmental goals than the middle-aged
adults, f(282.34) =
-8.41,
p < .001, and the middle-aged
adults, in turn, perceived their goals as "younger" in age timing
than the old adults, t( 309.55) = -6.07,p<
.001.
It is important
to note that the mean ratings for the age timing of goals—34.82
for the young, 41.56 for the middle aged, and 47.62 for the
old—did not reflect the mean actual age of the subjects (young:
29 years; middle aged: 48 years; old: 66 years) but departed in
the direction of relatively older estimates for the young adults
and relatively younger estimates for the middle-aged and, espe-
cially, old adults.
In addition, the difference between participants' actual age
and their normative age estimates for the developmental goals
was analyzed. A 3 (age) x 2 (east-west) x 2 (sex) X 4
(occupation) ANO\A was performed on these differences. The
main effect of age was significant, F(2, 409) = 298.7, p =
.001,
7/
2
- .594. The young adults perceived their goals to be
more typical of older ages than their own actual age (goal age
- actual age - 5.99). In contrast, the middle aged located their
goals at younger typical ages than their own actual age (goal
- actual = -6.25), young versus middle: r(287.16) = 15.64,
p < .001. The older adults expressed an even more extreme
difference between goal age and own actual age (M = —18.67),
middle aged versus old: f(292.10) = 11.88, p < .001.
Age timing for prime of
life.
A 3 (age) X 2 (east-west) X
2 (sex) X 4 (occupation) ANOVA yielded a significant main
effect for age, F(2, 462) = 88.89, p = .001, T)
2
= .278. The
ages perceived to be the prime of life by the young, middle-
aged, and old adults are shown in Figure 4, along with partici-
pants'
actual ages and the normative age of goals. As can be
seen in Figure 4, the young adults expected the prime of life at
an earlier time than the middle-aged adults, f(318.22) =
-8.03,
p < .001, and the middle-aged adults gave an earlier age esti-
mate than the old adults, r(318.27) =
-6.05,
p < .001 (young:
M - 30.47; middle aged: M = 40.08; old: M = 49.75).
To identify differences in the trajectories for normative ages
of goals and perceptions of the prime of life, a 3 (age) X 2
(east-west) X 2 (sex) X 4 (occupation) X 2 (goal age-prime
of life) ANOVA was performed on the estimates for normative
age of goal and perceptions of prime of life jointly, with goal
age versus prime of life as a within-subjects factor. The Goal
Age-Prime of Life X Age interaction was significant, F(2>
409) - 6.27, p = .01, rj
2
= .030. As can be seen in Figure 4,
age at the prime of life exhibited a steeper increase across age
groups than the normative age ascribed to goals. Specifically,
goal age for the young adults was higher than age at the prime
of life, f(136) = 4.29, p < .001.
Discussion
The present study aimed at investigating major characteristics
of developmental regulation in adults at different ages through-
out adulthood on the basis of the OPS model of developmental
regulation (J. Heckhausen & Schulz, 1993; Schulz & Heck-
hausen, 1996). The OPS model of developmental regulation
proposes that an individual's developmental regulation reflects
the specific constraints and opportunities encountered in a given
developmental ecology. Aging-related challenges are character-
ized by increasing and ultimately inescapable constraints and
Losses in the potential for attaining growth and in terms of
the risks of decline in functioning. This process is particularly
pronounced in advanced old age. It was, therefore, expected
that, at increasing age during adulthood, individuals typically
become more aware of these aging-related threats to primary
control. Such awareness of lowered growth potential and re-
duced control at least in some major areas of life (see also J.
Heckhausen & Baltes, 1991) should then give rise to changes
in preferred control strategies. Older adults should focus on
age-appropriate goals in their selective primary control striving
and resort more than younger adults to compensatory secondary
control strategies to protect their motivational resources for
long-term primary control.
The present study investigated age-related differences in two
kinds of control strategies: selective primary control and com-
pensatory secondary control. Selective primary control in old
age was expected to be shown in the selection of developmental
goals representing aging-relevant domains and developmental
tasks (e.g., health) at the expense of goals representing domains
more appropriate for younger age groups. Moreover, goal selec-
tion in older compared with younger adults should focus on the
avoidance of developmental losses rather than on the striving
for developmental gains. Compensatory secondary control in
older compared with younger adults was expected to be reflected
in four characteristics: a greater flexibility of goal adjustment,
a higher satisfaction with one's present life, a greater tendency
to identity with younger age groups, and the perception that the
prime of life is at relatively higher ages. This scheme of predic-
tions about developmental regulation in old age is supported by
the present study's findings, which are discussed in detail in the
following section.

DEVELOPMENTAL REGULATION
183
Awareness
of
Decreased Potential
for
Developmental
Growth
The older adults exhibited awareness
of
developmental
con-
straints
and
aging-related threats
in
reduced optimism about
the
probability
of
achieving their goals compared with
the
younger
adults.
The
latter finding converges with previous findings
(Lachman,
1991;
Nurmi, Pulliainen,
&
Salmela-Aro, 1992)
and,
at first glance, might seem hardly surprising. However,
it is far
from trivial when considering
the
ongoing debate about
the
functional status
of
optimism
in the
regulation
of
action
and
self.
The view that optimism about important outcomes
is a
funda-
mental adaptive characteristic
of the
human mind
has
become
widely accepted (Bandura, 1977,1992; Scheier
&
Carver,
1992;
Taylor, 1989; Taylor
&
Brown, 1988). However, there have also
been warnings
of the
potential negative consequences
of per-
ceived invulnerability,
for
instance,
in the
context
of
health
pro-
tective behavior (Schwarzer,
1992,
1994).
The
latter view
pro-
poses
a
more elaborate model, which takes into account both
the positive effect
of
risk awareness
and the
promdtive effect
of
optimistic action-outcome contingencies
on
appropriate health
behavior.
The
present study's findings suggest that more realistic
assessments
of
control potential might well
be
adaptive
not
only
with regard
to
health behaviors
but
also
in the
context
of
aging.
Apparently, older adults
do
acknowledge constraints
in
potential
for growth
and
increased risks
for
decline without suffering
deficits
and in
fact coming
out
higher
in
their life satisfaction.
Moreover, this risk awareness
did not
seem
to
prompt these
older adults
to
abandon those goals, which seemed more difficult
to obtain
(but
still obtainable). From both
a
theoretical
and
applied psychology point
of
view,
it is
important
to
note that
the older adults
did not
simply adapt their choice
of
goals
to
maintain
a
certain level
of
success expectancy
but
selected
or
continued
to
strive
for
goals, even
if
they held less promise
of
success. Thus, these adults demonstrated more concern with
important personal goals than with maintaining
a
sense
of
opti-
mism
and
control over their personal goals. This
may
serve
to illustrate
the
relative importance
of
primary control striving
compared with striving
for
subjective well-being
and a
sense
of
self-efficacy
itself.
A similar argument
can be
made
for the
older adults' lowered
perceptions
of
personal control with regard
to
attainment
of
personal developmental goals.
An
approach viewing optimism
as universally adaptive would
not do
justice
to the
findings.
Moreover,
the
finding
of
decreased perceptions
of
personal
con-
trol over personal developmental goals
in old age
also contri-
butes
to the
debate about stability versus decline
in
perceptions
of personal control during
old age.
Earlier work
by
Lachman
(1986) showed cross-age stability
in
perceptions
of
internal
control, whereas
her
more recent research uncovered decreases
in internal control perceptions
for
age-sensitive domains (Lach-
man, 1991). Consistently, Lachman
has
found age-related
in-
creases
in
perceived external control
in
aging-sensitive domains,
such
as
health
and
memory. Brandtstadter (1989) found decline
in personal control perceptions
in
earlier research, whereas
his
more recent work suggested substantial stability
in
perceived
internal control (Brandtstadter
&
Rothermund, 1994).
While
the
work
by
Lachman
and
Brandtstadter deals with
control beliefs about more general domains
of
life, Nurmi
and
his colleagues investigated control perceptions regarding
sub-
ject-specific life goals (Nurmi
et al., 1992)
similar
to the
goals
targeted
in the
present study.
On the
basis
of
unidimensional
bipolar measurement
of
internality
and
externality, Nurmi
et al.
(1992) found increasingly external control ascriptions
in
older
adults. These findings
are
consistent with Lachman's research
and, presuming that increases
in
external control ascriptions
may also imply decreases
in
perceived personal control, also
converge with
the
findings
of the
present study.
Selective Primary Control
The life span theory
of
control proposes that primary control
striving remains stable across
the
life span, even though
the
opportunities
for
primary control decrease
in old age
(Schulz
&
Heckhausen, 1996). This general prediction
was
confirmed
in
that
no age
differences were found
in
tenacious goal pursuit
across
the
adult life span. However, primary control striving
in
old age is
expected
to
address selectively age-appropriate
developmental tasks (Havighurst, 1952). This
age
gradedness
in selective primary control striving
is
discussed
in the
following
section.
The present study's findings
are
consistent with
the
assumption
that
the
older adults selectively focused their attempts
to
promote
primary control
on
age-appropriate goals
in
the domains of health,
community,
and
leisure activities.
It
should
be
noted that
a
contin-
uous increase from young
to
middle-aged adults
and
from middle-
aged
to old
adults
was
found only
for
health-related goals,
whereas leisure-related goals increased only from middle
to old
age
and
community-related goals only increased between
the
young
and the
middle-aged group, remaining stable thereafter.
This pattern
of
findings converges with common conceptions
of
developmental tasks (Havighurst,
1952) and
previous research
(e.g., Nurmi,
1992) in
that community engagement
is
viewed
as
a social responsibility
of
midlifers
and
people
in
advanced
age,
whereas free time
and
leisure gain ground only after retirement
in
old
age. During midlife,
in
contrast, health issues (particularly
preventive health care)
may
increasingly concern individuals
and
then become even more central
as an
actual decline
in
physical
functioning sets
in
during
old age.
Older adults disengaged from obsolete developmental goals
in
the
life domains
of
work, finances,
and
family, which provide
less primary control potential
in old age
(e.g., because
of
retire-
ment).
These patterns
of
decline
in
life goals concerning work,
family,
and
finances
are
consistent with
age
differences found
in previous research (Brandstadter & RDthermund, 1994; Nurmi,
1992;
Salmela-Aro, Nurmi, Aro, Poppius,
&
Riste, 1993). Such
goal disengagements presumably
are
facilitated
by the
increased
use
of
strategies
of
compensatory secondary control
(see Com-
pensatory Secondary Control section below).
The
present study
is limited
in
that
the
temporal
or
even causal relation between
disengagement from obsolete goals
and
engagement
in
age-ap-
propriate goals
is not
explicitly addressed. This should
be
done
by longitudinal research tracking such shifts around develop-
mental transitions.
Selective
and
age-adapted goal striving
in
older adults
is
also
reflected
in an
aging-related shift from goals directed
at
develop-
mental gains
to
goals targeted
at
avoiding losses. Striving
for

184
HECKHAUSEN
gains decreased continuously from early to middle adulthood
and from midlife to old age. In contrast, avoiding losses con-
stantly increased from young to middle and from middle to late
adulthood. This age-related shift from aiming for growth to
resorting to defense corresponds to previous findings about nor-
mative conceptions of development in adulthood held by adults
at different age levels. J. Heckhausen, Dixon, and Baltes (1989)
found an age-related shift in expected developmental change
from a predominance of gains in early adulthood to increasing
losses in old age. In keeping with such normative conceptions
about the increasing risk of losses, older adults more than
younger adults focus on actively avoiding age-related decline,
thereby demonstrating adaptation to age-graded structures of
opportunities and constraints (J. Heckhausen, in press).
Compensatory Secondary Control
The awareness of less fortunate prospects for developmental
change and personal control in old age as well as the necessity
to disengage from previously important personal goals because
of waning opportunities and ever increasing constraints presents
a substantial threat to the individual's control system. This could
seriously undermine the individual's motivational and emotional
resources for future, and still feasible, primary control striving.
Therefore, these losses in primary control need to be balanced
by increased use of strategies of compensatory secondary con-
trol. It was thus predicted that older adults would use more and
diverse compensatory secondary control strategies.
As expected and in accordance with previous research
(Brandtstadter & Renner, 1990; Brandtstadter et al., 1993), the
older adults demonstrated greater flexibility to adjust their goals.
One should keep in mind here that flexibility of goal adjustment
was measured as a generalized behavioral tendency rather than
in any goal-specific way. Thus, flexibility of goal adjustment
does not necessarily imply disengagement from a goal but also
includes lowering aspirations for a given goal. Thus, apart from
disengaging from obsolete goals, high flexibility might be re-
flected, for instance, in older adults sticking to their health-
related goals but lowering their aspirations from full functioning
to basic functioning in important activities of daily living (M.
M. Baltes, Mayr, Borchelt, Maas, & Wilms, 1993).
Another way of bringing oneself in line with inevitable reali-
ties—that is, compensatory secondary control—is to rescale
aspirations with regard to life satisfaction (e.g., Brandtstadter &
Rothermund, 1994). It was predicted that older adults would
express greater satisfaction with their present life than younger
adults. This prediction was confirmed overall. There were, how-
ever, some exceptions to this finding pertaining to occupational
groups with higher educational background. These exceptions
may be based on a greater scope for improvement even in old
age in adults at higher socioeconomic status, who command
greater social resources, and whose productivity may depend
more on abilities that fade less rapidly with age.
Moreover, compensatory secondary control is exemplified by
identifying with higher status social reference groups, so as to
elevate respective self-images and thus moderate negative effects
of losses in control on self-esteem. In the context of aging, we
expected that older adults would show a greater tendency to
identify with younger age groups. Whereas young adults identi-
fied with their peers, middle-aged adults and especially old adults
estimated various aspects of subjective age as younger than their
actual age. This pattern of findings accords with previous research
on subjective age identification (Filipp & Ferring, 1989; J. Heck-
hausen & Krueger, 1993; Montepare & Lachman, 1989).
In accordance with the findings on subjective age, the norma-
tive age timing of developmental goals also deviated from parti-
cipants' actual age. Older adults perceived their own develop-
mental goals to be substantially "younger" than themselves,
locating them in midlife rather than in old age. Middle-aged
adults were more oriented toward younger midlifers than them-
selves. Conversely, younger adults perceived their own goals to
be typical of somewhat older individuals. A similar pattern of
findings was obtained in a previous study on self-other differ-
ences in images of aging (J. Heckhausen & Krueger, 1993).
Apparently, adults at different ages view their developmental
aspirations as indicative of a reference age group that is some-
what higher in social status and still not too far apart from one's
own age. Thus, to put it in an illustrative manner, younger adults
are oriented toward somewhat more mature individuals, whereas
middle-aged and especially old adults aspire to less advanced
aging. It is important to note that the older adults did not simply
pick the same age referent for their developmental goals as the
younger age groups (see Figure 4), which one would expect if
sheer wishful thinking was determining participants' develop-
mental aspirations. Instead, neighboring age groups are selected
as reference for their developmental goals, thus rendering the
difficulty of the developmental aspiration intermediate and,
therefore, reasonable. Selecting intermediate difficulty is known
to be ideal for the promotion of mastery (Atkinson, 1957; H.
Heckhausen, 1991).
Together, the findings on subjective age and on normative
age timing of one's own goals reflect concerted strategies of
compensatory secondary control to construct a positively biased
view of the self with regard to age, the key dimension associated
with feared losses. In addition, positively biased conceptions
of the self with regard to age may also reflect upward social
comparisons that promote the struggle for primary control even
for those goals that are typically perceived as more difficult to
obtain in old age.
A complementary secondary control strategy for reinterpret-
ing the relation between the self and chronological age was
found regarding perceptions about the prime of life. The older
the subject, the higher the age considered to be the prime of
life.
Although the normative age ascribed to goals presumably
reflected adults
1
upward developmental aspirations, estimated
prime of life may represent a reinterpretation of life course
phases in view of and adapted to one's own actual age. Thus,
one may speculate that altering one's view of the prime of
life (relative to actual age) primarily serves to compensate for
experiences or feared losses by means of secondary control
without directly feeding into the struggle for primary control as
in the case of subjective age identification and normative age
timing of goals. In this context, it may be interesting to consider
the fact that young adults viewed their own age as the prime of
life but were more oriented toward midlife in terms of develop-
mental goals. Thus, early adulthood might be viewed as highly
desirable but not as optimal in terms of promoting major devel-
opmental goals of adulthood.

DEVELOPMENTAL REGULATION
185
Summary and Conclusion
The present study's findings on age-related differences in
developmental regulation confirm the predictions based on the
OPS life span model (J. Heckhausen & Schulz, 1993). It is
important to note that all of the age differences found in this
study were present across a wide range of participant character-
istics in terms of gender, occupational status, and East versus
West Berlin origin. The pattern of age-related differences found
in this study suggests several functional characteristics of devel-
opmental regulation in old age, discussed next.
Aging-related losses are not denied but are acknowledged by
older adults in various ways. Older adults tend to be less opti-
mistic about the likelihood of attaining their goals and the degree
to which they can control goal attainment. However, this likely
realistic assessment of constraints to future development does
not discourage the aged from striving for developmental goals
that are still obtainable. In fact, older adults seem to bank on
selective primary control for age-appropriate goals at the ex-
pense of other obsolete goals. Selective primary control is partic-
ularly focused on avoiding impending loss. Although striving to
realize the developmental potential still available, older adults
also come to terms with unavoidable losses without falling prey
to despair. This is achieved by various strategies of compensa-
tory secondary control. First, older individuals disengage from
obsolete goals that are no longer obtainable because of irrevers-
ible changes in social roles related to work, family status, and
biological decline. Moreover, older adults withdraw from striv-
ing for developmental gains in favor of focusing on loss avoid-
ance.
Given the constrained potential for future improvement,
it is adaptive that older adults feel more satisfied with the present
than younger adults. Moreover, older adults use secondary com-
pensatory control to moderate their age status. Older adults
typically identify with middle adulthood and perceive their own
developmental goals as typical for middle-aged adults. In this
way, older adults view themselves as exceptional cases among
the elderly and thereby probably boost their self-esteem. Finally,
they have their own adaptive outlook on the prime of life in
that they locate it closer to their own age than to early adulthood.
It may be interesting to consider briefly the pattern of control
strategies differentiating East from West Berliners because it
may represent an adaptation to a different scenario of develop-
mental challenges presented by radical sociohistorical change.
Overall, the study's findings suggest that East Berliners' com-
pared with West Berliners' developmental regulation reflects
heightened selective primary and selective secondary control
(J. Heckhausen, 1994). East Berliners focused more on work-
related developmental goals (i.e., goals related to economic sur-
vival).
Moreover, East Berliners were less engaged in goals
related to health, thus exhibiting selective investment in work-
related goals at the expense of other important goals. Enhanced
selective primary control in East Berliners was also reflected in
greater tenaciousness of goal pursuit. Finally, selective second-
ary control is reflected in the current lower life satisfaction
reported by East Berliners compared with West Berliners. In
this way, East Berliners probably felt a greater need and stronger
motivation to change their current life situation. This pattern of
combined selective primary and secondary control strategies
should be adaptive under conditions of radical sociohistorical
change (see also the concept of "control cycle" by Elder &
Caspi, 1990) because such societal transformations can be ex-
pected to increase the chances for gains and the risks of losses
(Elder, 1974). Thus, for better or worse, the impact of active
attempts to control one's life and development is enhanced dur-
ing major societal transformations.
In sum, human adults command a rich array of primary and
secondary control strategies to master selectivity and compensa-
tion in developmental regulation. These strategies allow the indi-
vidual to effectively pursue developmental goals within reach
and to abandon unobtainable and thus dysfunctional goals. In
this way, individuals can adapt their behavior specifically to a
given developmental ecology, take into account its constraints,
and make use of the opportunities for control. To determine
both the potential and limits in human developmental regulation,
further research should investigate other developmental scenar-
ios varying in degree of controllability, risks for failure, time
constraints in terms of "developmental deadlines" (J. Heck-
hausen, in press), payoffs for success, and negative versus posi-
tive trade-offs in multiple developmental goals.
References
Atkinson, J, W. (1957). Motivational determinants of risk-taking behav-
ior. Psychological Review, 64, 359-372.
Baltes, M. M., Mayr, U., Borchelt, M, Maas, I., & Wilms, U. (1993).
Everyday competence in old and very old age: An interdisciplinary
perspective. Ageing and Society, 13, 657-680.
Baltes, P. B., & Baltes, M. M. (1990). Psychological perspectives on
successful aging: The model of selective optimization with compensa-
tion. In P. B. Baltes & M. M. Baltes (Eds.), Successful aging: Per-
spectives from the behavioral sciences (pp.
1-34).
New ^ork: Cam-
bridge University Press.
Baltes, P. B., Lindenberger, U., & Staudinger, U. M. (in press). Life
span theory in developmental psychology. In
W.
Damon (Series Ed.) &
R. M. Lerner (Vol. Ed.), Handbook of child psychology: Vol. 1, Theo-
retical model of human development. New \brk: Wiley.
Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behav-
ioral change. Psychological Review, 84, 191-215.
Bandura, A. (1982). Self-efficacy mechanisms in human agency. Ameri-
can Psychologist, 37, 122-147.
Bandura, A. (1992). Exercise of personal agency through the self-effi-
cacy mechanism. In R. Schwarzer (Ed.), Self-efficacy: Thought con-
trol of action (pp. 3-38). Washington, DC: Hemisphere.
Brandtstadter, J. (1989). Personal self-regulation of development: Cross-
sequential analyses of development-related control beliefs and emo-
tions.
Developmental Psychology, 25, 96-108.
Brandtstadter, J. (in press). Action perspective in human development.
In W. Damon (Series Ed.) & R. M. Lerner (Vol. Ed.), Handbook of
child psychology: Vol. 1. Theoretical models of human development.
New \brk: Wiley.
Brandtstadter, J., & Renner, G. (1990). Tenacious goal persuit and flexi-
ble goal adjustment: Explication and age-related analysis of assimila-
tive and accommodative strategies of coping. Psychology and Aging,
5, 58-67.
Brandtstadter, J., & Rothermund, K. (1994). Self-percepts of control
in middle and later adulthood: Buffering losses by rescaling goals.
Psychology and Aging, 9, 265-273.
Brandtstadter, J., Wentura, D., & Greve, W. (1993). Adaptive resources
of the aging
self:
Outlines of an emergent perspective. International
Journal of Behavioral Development, 16, 323-349.
Cantor, N., & Fleeson, W. W. (1991). Life tasks and self-regulatory
processes. In M. L. Maehr & P. R. Pintrich (Eds.), Advances in moti-

186
HECKHAUSEN
vation and achievement (Vol. 7, pp. 327-369). Greenwich, CT JAI
Press.
Cantor, N., & Fleeson, W. W. (1994). Social intelligence and intelligent
goal pursuit: A cognitive slice of motivation. In G. Spaulding (Ed.),
Nebraska symposium on motivation (Vol. 41, pp. 125-179). Lincoln:
University of Nebraska Press.
Cantor, N., Norem, J. K., Niedenthal, P. M., Langston, C. A., & Brower,
A. M. (1987). Life tasks, self-concept ideals, and cognitive strategies
in a life transition. Journal of Personality and Social Psychology, 53,
1178-1191.
Carstensen, L. L. (1993). Motivation for social contact across the life
span. A theory of socioemotional selectivity. In J. Jacobs (Ed.), Ne-
braska symposium on motivation: Developmental perspectives on mo-
tivation (Vol. 40, pp. 209-254). Lincoln: University of Nebraska
Press.
Elder, G. H., Jr. (1974). Children of the Great Depression. Chicago:
University of Chicago Press.
Elder, G. H., Jr., & Caspi, A. (1990). Studying lives in a changing
society: Sociological and personologkal explorations. In A. Rabin, R.
Zucker, R. Emmons, & S. Frank (Eds.), Studying persons and lives
(pp.
201-247). New York: Springer.
Emmons, R. A. (1986). Personal strivings: An approach to personality
and subjective well-being. Journal of Personality and Social Psychol-
ogy, 51, 1058-1068.
Emmons, R. A., & Diener, E. (1986). A goal-affect analysis of everyday
situational choices. Journal of Research in Personality, 20, 309-326.
Filipp,
S.-H.,
& Ferring, D. (1989). Zur Altes- und Bereichsspezifitat
subjektiven Alterserlebens [Age and domain-specificity in subjective
age experience]. Zeitschriftfur Entwicklungspsychologie undPdda-
gogische Psychologie, 21, 279-293.
Gollwitzer, P. M. (1987). Suchen, Finden und Festigen der eigenen Iden-
titat: Unstillbare Zielintentionen
[
Searching, rinding, and consolidat-
ing one's own identity: Insatiable goal intentions]. In H. Heckhausen,
P.
M. Gollwitzer, & F. E. Weinert (Eds.), Jenseits des Rubikon: Der
Wille in den Humanwissenschaften [On the other side of die rubicon:
The will in the humanities] (pp. 176-189). Berlin: Springer-Verlag.
Gollwitzer, P. M., & Kinney, R. F. (1989). Effects of deliberative and
implemental mind-sets on illusion of control. Journal of Personality
and Social Psychology, 56, 531-542.
Harris, L., & Associates. (1981). Aging in the Eighties: America in
transition. Washington, DC: National Council on the Aging.
Havighurst, R. J. (1952). Developmental tasks and education. New
York: McKay.
Heckhausen, H. (1991). Motivation and action. New "fork: Springer.
Heckhausen, J. (1994). Entwicklungsziele und Kontrolliiberzeugungen
Ost- und Westberliner Erwachsener [Developmental goals and control
beliefs in East and West
Berliners].
In G. Trommsdorff (Ed.), Psycho-
logische Aspekte des sozio-politischen Wandels in Ostdeutschland
[Psychological aspects of sociopolitical change in East Germany] (pp.
124-133). Berlin: de Gruyter.
Heckhausen, J. (in press). Developmental regulation in
adulthood.
New
York: Cambridge University Press.
Heckhausen, J., & Baltes, P. B. (1991). Perceived controllability of
expected psychological change across adulthood and old age. Journal
of Gerontology: Psychological Sciences, 46, 165-173.
Heckhausen, J., Dixon, R. A., & Baltes, P. B. (1989). Gains and losses
in development throughout adulthood as perceived by different adult
age groups. Developmental Psychology, 25,
109-121.
Heckhausen, J., & Fleeson, W. (1993). Primary-secondary control shifts
in developmental regulation around age deadlines for developmental
tasks: Actional and post-actional functioning with impending versus
passed action opportunities. Unpublished manuscript, Max Planck
Institute for Human Development and Education, Berlin.
Heckhausen, J., & Hundertmark, J. (1995). Action-related beliefs in
adulthood:
Measurement of perceived agency, means-ends relations,
and personal control in three adult life domains. Unpublished manu-
script, Max Planck Institute for Human Development and Education,
Berlin.
Heckhausen, J., & Krueger, J. (1993). Developmental expectations for
the self and most other people: Age grading in three functions of
social comparison. Developmental Psychology, 29, 539-548.
Heckhausen, J., & Schulz, R. (1993). Optimization by selection and
compensation: Balancing primary and secondary control in life-span
development. International Journal of Behavioral Development, 16,
287-303.
Heckhausen, J., & Schulz, R. (1994). Primacy of primary control as a
universal feature of human behavior. Paper presented at 13th Biennial
Meeting of the International Society for the Study of Behavioral De-
velopment, Amsterdam.
Heckhausen, J., & Schulz, R. (1995). A life-span theory of control.
Psychological Review, 102, 284-304.
Kastenbaum, R., Derbin, V, Sabatini, P., & Artt, S. (1981). "The ages
of me*': Toward personal and interpersonal definitions of functional
aging. In Robert Kastenbaum (Ed.), Old age and the new scene (pp.
48-67).
New York: Springer.
Kliegl, R., Smith, J., & Baltes, P. B. (1990). On the locus and process
of magnification of age differences during mnemonic training. Devel-
opmental Psychology, 26, 894-904.
Klinger, E. (1975). Consequences of commitment and disengagement
from incentives. Psychological Review, 82, 1-25.
Klinger, E., Barta, S. G., & Maxeiner, M. E. (1980). Motivational corre-
lates of thought content frequency and commitment. Journal of Per-
sonality and Social Psychology, 39, 1222-1237.
Lachman, M. E. (1986). Locus of control in aging research: A case for
multidimensional and domain-specific assessment. Psychology and
Aging, 1, 34-40.
Lachman, M. E. (1991). Perceived control over memory aging: Devel-
opmental and intervention perspectives. Journal of Social Issues, 47,
159-175.
Lerner, R. M., & Busch-Rossnagel, N. A. (1981). Individuals as produc-
ers of their development: Conceptual and empirical bases. In R. M.
Lerner & N. A. Busch-Rossnagel (Eds.), Individuals as producers of
their development: A life-span perspective (pp.
1-36).
New \brk:
Academic Press.
Little, B. R. (1983). Personal projects: A rationale and method for inves-
tigation. Environment and Behavior, 15, 273-309.
Marsiske, M., Lang, F. R., Baltes, P. B., & Baltes, M. M. (1995). Selec-
tive optimization with compensation: Life-span perspectives on suc-
cessful human development. In R. A. Dixon & L. Backman (Eds.),
Psychological compensation: Managing losses and promoting gains.
(pp.
35-79). Mahwah, NJ: Erlbaum.
Montepare, J. M., & Lachman, M. E. (1989). "You're only as old as
you feel": Self-perceptions of age, fears of aging, and life satisfaction
from adolescence to old age. Psychology and Aging, 4, 73-78.
Nelson, E. A., & Dannefer, D. (1992). Aged heterogeneity: Fact or
fiction? The fate of diversity in gerontological research. The Gerontol-
ogist,
32, 17-32.
Nurmi, J.-E. (1992). Age differences in adult life goals, concerns, and
their temporal extension: A life-course approach to future-oriented
motivation. International Journal of Behavioral Development, 15,
487-508.
Nurmi, J.-E., Pulliainen, H., & Salmela-Aro, K. (1992). Age differences
in adults' control beliefs related to life goals and concerns. Psychology
and Aging, 7, 194-196.
Palys,
T. S., & Little, B. R. (1983). Perceived life satisfaction and the
organization of personal project systems. Journal of Personality and
Social Psychology, 44, 1221-1230.
Rothbaum, F., Weisz, J. R., & Snyder, S. S. (1982). Changing the world

DEVELOPMENTAL REGULATION
187
and changing the
self:
A two-process model of perceived control.
Journal of Personality and Social Psychology, 42, 5-37.
Salmela-Aro, K., Nurmi, J.-E., Aro, A., Poppius, E., & Riste, J. (1993).
Age differences in adults' personal projects. Journal of Social Psy-
chology,
133,
415-417.
Scheier, M. K, & Carver, C. S. (1992). Effects of optimism on psycho-
logical and physical well-being: Theoretical overview and empirical
update. Cognitive Therapy and Research, 16, 201-228.
Schulz, R., & Heckhausen, J. (1996). A life span model of successful
aging. American Psychologist, 51, 702-714.
Schulz, R., Heckhausen, J., & Locher, J. (1991). Adult development,
control, and adaptive functioning. Journal of Social Issues, 47, 177-
196.
Schwarzec R. (1992). Self-efficacy in the adoption and maintenance of
health behaviors: Theoretical approaches and a new model. In R.
Schwarzer (Ed.), Self-efficacy: Thought control of action (pp. 217-
242).
Washington, DC: Hemisphere.
Schwarzer, R. (1994). Optimism, vulnerability, and self-beliefs as
health-related cognitions. A systematic overview. Psychology and
Health, 9, 161-180.
Taylor, S. E. (1989). Positive illusions: Creative self-deception and the
healthy
mind.
New \brk: Basic Books.
Taylor, S. E., & Brown, J. D. (1988). Illusion and well-being: A social
psychological perspective on mental health. Psychological Bulletin,
103,
193-210.
Taylor, S.E., & Lobel, M. (1989). Social comparison activity under
threat: Downward evaluation and upward contacts. Psychological Re-
view, 96, 569-575.
Thompson, L. A. (1995). Encoding and memory for visible speech and
gestures: A comparison between young and older adults. Psychology
and
Aging,
10, 215-228.
Wadsworth, M., & Ford, D. H. (1983). Assessment of personal goal
hierarchies. Journal of Counseling Psychology, 30, 514-526.
Wills,
T.
A. (1981). Downward comparison principles in social psychol-
ogy. Psychological Bulletin, 90,
245-271.
Wood, J. V. (1989). Theory and research concerning social comparison
of personal attributes. Psychological Bulletin, 106, 231-248.
Received September 15, 1994
Revision received February 13, 1996
Accepted February 13, 1996 •