Journal of Behavioral Medicine, Vol. 29, No.

3, June 2006 (

C 2006)

DOI: 10.1007/s10865-005-9044-1

The Life Engagement Test: Assessing Purpose in Life

Michael F. Scheier,1,5 Carsten Wrosch,2 Andrew Baum,3 Sheldon Cohen,4 Lynn M. Martire,3
Karen A. Matthews,3 Richard Schulz,3 and Bozena Zdaniuk3

Accepted for publication: December 01, 2005

Published online: March 24, 2006

This article describes a 6-item scale, the Life Engagement Test, designed to measure purpose
in life, defined in terms of the extent to which a person engages in activities that are person-
ally valued. Psychometric data are presented including information about the scale’s factor
structure, internal consistency, test-retest reliability, convergent validity, discriminant predic-
tive validity, and norms. The data suggest that the Life Engagement Test is psychometrically
sound across different gender, age, and ethnic groups and is appropriate for wider use. Dis-
cussion centers on the use of the Life Engagement Test in behavioral medicine and health
psychology research and recent associations that have begun to emerge between the scale
and health-relevant outcomes.
KEYWORDS: purpose in life; life engagement; assessment; factor structure; psychometrics.

Death has been defined as the absence of behav- provide the mechanism by which a person remains
ior (Kaplan, 1990). We agree that behavior is impor- behaviorally engaged in life. According to this view,
tant for health and that it does go hand-in-hand with behavior occurs either because the behavior repre-
being alive. Some have even suggested that it is en- sents a valued goal in and of itself (e.g., exercising for
gagement in behavior that sustains life (Carver and exercise sake) or because it is instrumental in achiev-
Scheier, 1998). Given the central role that behavior ing a more abstract, higher order goal that is valued
plays in living, it is important to ask why people act. (e.g., exercising in order to be “healthy”).
What is it that causes people to behave and remain This brief report describes a new scale designed
engaged in what they do? to measure purpose in life. It is our belief that pur-
Recent models of behavioral self-regulation pose in life represents an important but overlooked
(Carver and Scheier, 1981, 1990, 1998), them- psychosocial predictor of health outcomes. Life is
selves descendents of generations of expectancy- full of situations in which desired activities must be
value models of motivation (Atkinson, 1964; Vroom, abandoned—e.g., the person who still loves to work,
1964; Feather, 1982; Shah and Higgins, 1997), suggest but is forced to retire because of age; the carpenter
that two elements are important in creating behavior: who gets great satisfaction working with wood, but
(a) the ability to identify goals that are valued and whose arthritis is so severe that he or she can no
(b) the perception that those goals are attainable. Of longer grip the required tools. In these contexts, the
these two elements, it is the value dimension that is person must find alternative meaningful activities in
of interest here. Valued goals are important because which to engage. If new activities are not found, the
they provide a purpose for living. Valued goals also person’s life feels empty and is without purpose (cf.
Wrosch et al., 2003a). We believe that feeling states
1 Department of Psychology, Carnegie Mellon University, Pitts- such as these will have important adverse effects on
burgh, Pennsylvania. psychological and physical well-being (Wrosch et al.,
2 Concordia University, Quebec.
3 University of Pittsburgh, Pennsylvania.
2003b).
4 Carnegie Mellon University, Pennsylvania. Given that purpose in life might be useful to
5 To whom correspondence should be addressed e-mail: scheier@ measure, why do we need a new scale? Are there
cmu.edu. not existing scales already available that would be

291
0160-7715/06/0600-0291/0

C 2006 Springer Science+Business Media, Inc.
292 Scheier et al.

suitable for use? There are at least two reasons why of the Pittsburgh Mind-Body Center. Sample 1 con-
a new scale is needed. First, existing scales are of- sisted of community-dwelling men and women par-
ten time insensitive. For example, they might ask ticipating in a study on infectious disease. Samples
whether the person’s life has been filled with purpose 2 and 3 consisted of female osteoarthritis patients
rather than asking whether the person is currently ex- and their male spouses, respectively, participating in
periencing purpose in life. This makes it difficult to a study of psychosocial factors in adjustment to os-
assess changes in purpose in life over time, which one teoarthritis. Sample 4 consisted of a group of women
might want to do, for example, in order to monitor participating in a study examining changes in cardio-
changes that occur as someone progresses through vascular risk factors as a function of undergoing tran-
the later stages of a degenerative or fatal disease. sition through the menopause. Samples 5 and 6 con-
Second, other scales often contain items that sisted of early and late stage breast cancer patients,
measure constructs in addition to purpose in life; e.g., respectively, participating in a study of adjustment to
items that measure life satisfaction, contentment, or breast cancer. Samples 7 and 8 consisted of college
meaning. Such confounding makes it difficult to de- undergraduates. Table II presents basic demographic
termine which components contained within a scale data for these various samples.
are responsible for producing any associations that
emerge. In this regard, construction of the current
Factor Structure
scale explicitly grew out of the theoretical framework
provided by contemporary models of behavioral self-
To examine the factor structure of the LET, we
regulation. Because of this focus, we call our scale
conducted exploratory factor analyses across the dif-
the Life Engagement Test, or LET. The intent of the
ferent samples (using principal component analyses
scale is to provide an index of purpose in life by as-
with varimax rotation and Kaiser normalization, re-
sessing the extent to which a person considers his or
taining factors with Eigen values >1). The results of
her activities to be valuable and important.
these factor analyses revealed a one-factor solution
for all samples, which accounted for between 43%
METHODS AND RESULTS
and 62% of the variance among the items. The factor
loadings of the six items of the LET are reported in
Scale Format, Instructions for Administration,
Table I. As can be seen, we obtained high factor load-
and Scoring
ings for all of the items of the LET, ranging across
samples from .57 to .86, averaging .71.
The LET consists of six items (see Table I), three
items framed in a positive direction (Items 2, 4, and 6)
and three items framed in a negative direction (Items Scale Norms and Internal Consistency
1, 3, and 5). The following instructions are used to ad-
minister the scale: “Please answer the following ques- Table III displays the means, standard devia-
tions about yourself by indicating the extent of your tions, and reliability coefficients of the LET across
agreement using the following scale: 1 = strongly dis- the eight samples. We obtained acceptable Cron-
agree; 2 = disagree; 3 = neutral; 4 = agree; 5 = bach’s alphas in all cases, ranging between .72 and
strongly agree. Be as honest as you can throughout, .87, averaging .80. Although the means across the
and try not to let your response to one question influ- samples were close in value, given the sample sizes,
ence your response to other questions. There are no the variation among the sample means was signif-
right or wrong answers.” icant, F(7, 2244) = 6.17, p<.001. Follow-up Tukey
The LET is scored in two steps. First, Items 1, 3, tests revealed that Sample 7 was significantly lower
and 5 are reverse coded (5 = 1, 4 = 2, 3 = 3, 2 = 4, than Samples 1, 4, and 5. (The rest of the sample
and 1 = 5). Second, the six items are summed. means did not differ.)

Samples Used to Identify Psychometric Properties Gender, Age, and Ethnic Differences

Eight different samples were used to establish To determine if there were any differences in
the psychometric properties of the LET. The first the psychometric properties of the LET as a func-
six samples were drawn from persons participating tion of gender, age, or ethnicity, we reevaluated
in one of the four main projects of the first phase the scale’s factor structure, norms, and reliability by
The Life Engagement Test: Assessing Purpose in Life 293

Table I. The Life Engagement Test: Items and Factor Loadings

Sample

Items of the life engagement test 1 2 3 4 5 6 7 8

1. There is not enough purpose in .76 .81 .72 .73 .78 .79 .74 .59
my life. (R)a
2. To me, the things I do are all .67 .67 .72 .70 .77 .85 .67 .60
worthwhile
3. Most of what I do seems trivial .66 .78 .81 .80 .86 .90 .80 .70
and unimportant to me. (R)a
4. I value my activities a lot .66 .62 .73 .71 .81 .69 .69 .64
5. I don’t care very much about the .59 .72 .67 .76 .79 .85 .69 .71
things I do. (R)a
6. I have lots of reasons for living .57 .70 .58 .56 .69 .61 .63 .65
Note. Sample 1 = community-based sample of younger adults (n=193).
Sample 2 = female osteoarthritis patients (n=183).
Sample 3 = male spouses of osteoarthritis patients (n=168).
Sample 4 = community-based sample of middle-aged women (n=378).
Sample 5 = women with early stage breast cancer (n=198).
Sample 6 = women with late stage breast cancer (n=86).
Sample 7 = undergraduate students (n=359).
Sample 8 = undergraduate students (n=511).
a Negatively formulated items were reverse scored.

stratifying across all samples (as the availability of ipants scored significantly higher on the LET than
identifying data allowed) first by gender, then by age Caucasians (mean = 25.5 versus mean = 24.7, re-
(trichotomized into tertiles), and finally by ethnicity spectively, t(1296) = −2.80, p<.01), but the abso-
(Caucasian versus African American, the only ethnic lute amount of the difference in scores was relatively
group for which a meaningful amount of data were small (.84).
available). The same one-factor solution emerged in
all of the subanalyses conducted. Cronbach’s alphas Test-Retest Reliability
were also acceptable in all subgroups (ranging from
.73 to .83). There were no mean differences as a To determine the test-retest reliability of the
function of gender or age. African American partic- LET, the women in Samples 5 (N = 178) and 6

Table II. Characteristics of the Samples

Sample

1 2 3 4 5 6 7 8
% Female 50.8 100 0 100 100 100 40.6 38.5
Average age in years 37 68 70 65 51 50 naa na
Age range in years 21—55 49–86 47–90 60–69 26–78 27–69 na na
% Married or living in 48.2 92.3 100 71.09 63.10 66.3 na na
marriage like relationship
% Caucasian 56.0 84.2 92.3 93.1 87.9 87.2 na na
% African American 37.3 13.1 5.4 5.8 11.1 8.1 na na
Note. Sample 1 = community-based sample of younger adults (n=193).
Sample 2 = female osteoarthritis patients (n=183).
Sample 3 = male spouses of osteoarthritis patients (n=168).
Sample 4 = community-based sample of middle-aged women (n=378).
Sample 5 = women with early stage breast cancer (n=198).
Sample 6 = women with late stage breast cancer (n=86).
Sample 7 = undergraduate students (n=359).
Sample 8 = undergraduate students (n=511).
a na = not available.
294 Scheier et al.

Table III. The Life Engagement Test: Means, Standard Deviations (SD), and Cronbach’s Alphas
Sample

1 2 3 4 5 6 7 8
Mean (SD) 25.1 (3.6) 24.3 (3.9) 24.8 (3.4) 24.9 (3.4) 25.3 (3.7) 24.1 (4.0) 23.6 (3.7) 24.4 (2.8)
Cronbach& alpha .73 .81 .80 .80 .87 .87 .79 .72
Note. Sample 1 = community-based sample of younger adults (n = 193).
Sample 2 = female osteoarthritis patients (n = 183).
Sample 3 = male spouses of osteoarthritis patients (n = 168).
Sample 4 = community-based sample of middle-aged women (n = 378).
Sample 5 = women with early stage breast cancer (n = 198).
Sample 6 = women with late stage breast cancer (n = 86).
Sample 7 = undergraduate students (n = 359).
Sample 8 = undergraduate students (n = 511).

(N = 62) and a subset of the men and women in Sam- item reduction based primarily on prior psychomet-
ples 7 (N = 55) and 8 (N = 61) were administered the ric analyses (e.g., Barefoot et al., 1989; for discussion
LET twice, approximately 4 months apart. The test- of the general validity of using abbreviated scales, see
retest correlations ranged from .61 to .76, suggesting Shrout and Yager, 1989).
that the LET is moderately stable, at least over the Table IV presents the correlations between the
period of several months. LET and these different measures across the eight
samples examined (note that not all samples got
Convergent Validity all measures, which explains why some of the cells
have missing values). As can be seen in Table IV,
To examine the convergent validity of the LET, the LET was widely correlated with the psychoso-
we correlated the LET with a variety of other psy- cial and health-relevant variables that were assessed,
chosocial measures, many of which have been associ- in intuitively plausible directions. For example, sig-
ated with health-relevant outcomes in prior research. nificant positive associations emerged between the
We also correlated the LET with several health- LET and optimism, life satisfaction, general health,
relevant variables directly. The attributes measured and self-esteem, and significant negative correlations
included perceived stress (Cohen et al., 1983), self- emerged between the LET and perceived stress, hos-
mastery (Pearlin and Schooler, 1978), depressive tile attitudes, and depression. Moreover, the mag-
symptoms [as assessed by either a 10-item abbre- nitude of the correlations suggest that the LET is
viated version (Andresen et al., 1994) of the CES- related to these other factors, but not so highly
D (Radloff, 1977) or by the Brief Symptom Inven- related as to suggest that the constructs are the
tory (Derogatis and Melisaratos, 1983)], optimism same.
(Scheier et al., 1994), life satisfaction (Diener et al., We explicitly acknowledge that some of the cor-
1985), self-esteem (Rosenberg, 1965), hostile affect relations in Table IV are low; e.g., the correlations
(Cook and Medley, 1954), anger-in expression style with reports of pain and sleep efficiency are some-
(Spielberger et al., 1985), emotional stability, ex- what lower than what might be expected. It is un-
traversion, agreeableness, openness to experience, clear why some of these correlations are low, but it
conscientiousness (all from Goldberg, 1992), general may be that purpose in life has less of an impact on
health, health-related physical and mental function- these particular aspects of functioning than it does on
ing (all from Ware and Sherbourne, 1992), marital others. A definitive answer to this question, however,
adjustment (Locke and Wallace, 1959), anxiety, so- will have to await further research.
maticism and hostility from the Brief Symptom In-
ventory (Derogatis and Melisaratos, 1983), percep- Discriminant Predictive Validity
tion of sleep efficiency (Buysse et al., 1989), social
network diversity (Cohen et al., 1997), social network The LET was specifically designed to assess pur-
size (Cohen et al., 1997), and perceptions of social pose in life by assessing the extent to which peo-
support (Cohen et al., 1985). Abbreviated versions ple engaged in activities that they found valuable
of some of the assessment instruments were used and significant. We focused the LET in this way be-
to make the overall protocol length tolerable, with cause we believe it is this aspect of purpose in life
The Life Engagement Test: Assessing Purpose in Life 295

Table IV. Convergent Validity: Correlations Between the Life Engagement Test and the Other Psychosocial Factors
Correlation with the Life Engagement Test

Sample

1 2 3 4 5 6 7 8
1. Optimism .39∗∗ .43∗∗ .41∗∗ .54∗∗ .61∗∗ .51∗∗ .54∗∗ .48∗∗
2. Self-mastery .52∗∗ .55∗∗ .43∗∗ .46∗∗ .53∗∗ .63∗∗ — —
3. Self-esteem .44∗∗ .48∗∗ .43∗∗ .48∗∗ .61∗∗ .53∗∗ — —
4. Hostile affect −.33∗∗ −.22∗∗ −.33∗∗ −.17∗∗ −.36∗∗ −.21 — —
5. Anger-in −.32∗∗ −.42∗∗ −.40∗∗ −.26∗∗ −.29∗∗ −.39∗∗ — —
6. Emotional stability .34∗∗ .40∗∗ .36∗∗ .24∗∗ .48∗∗ .30∗∗ .28∗∗ —
7. Extraversion .25∗∗ .20∗∗ .19∗ .24∗∗ .31∗∗ .48∗∗ .31∗∗ —
8. Agreeableness .32∗∗ .28∗∗ .43∗∗ .09 .34∗∗ .25∗ .30∗∗ —
9. Openness .23∗∗ .21∗∗ .24∗∗ .29∗∗ .29∗∗ .41∗∗ .39∗∗ —
10. Conscientiousness .32∗∗ .34∗∗ .30∗∗ .17∗∗ .31∗∗ .12 .32∗∗ —
11. Physical Functioning — — — .07 .18∗∗ .32∗∗ — —
12. Social Functioning — — — .47∗∗ .26∗∗ .28∗∗ — —
13. Role Disruption — — — — .24∗∗ .15∗ .31∗∗ — —
Physical Health
14. Role Disruption — — — — .19∗∗ .35∗∗ .42∗∗ — —
Emotional Health
15. Mental Health — — — .32∗∗ .49∗∗ .44∗∗ — —
16. Vitality — — — .43∗∗ .32∗∗ .34∗∗ — —
17. Pain — — — .29∗∗ .12 .21∗ — —
18. General Health — — — .34∗∗ .38∗∗ .30∗∗ — —
19. Social support .40∗∗ .39∗∗ .53∗∗ .39∗∗ .50∗∗ .46∗∗ — —
20. Social network size .32∗∗ .37∗∗ .30∗∗ .26∗∗ .29∗∗ .22∗ — —
21. Social network diversity .27∗∗ .29∗∗ .27∗∗ .18∗∗ .31∗∗ .24∗ — —
22. Marital adjustment .26∗∗ .25∗∗ .40∗∗ .28∗∗ .48∗∗ .35∗∗ — —
23. Depression (CES-D −.33∗∗ −.49∗∗ −.45∗∗ −.49∗∗ −.42∗∗ −.47∗∗ — —
derived)
24. Perceived stress −.44∗∗ −.51∗∗ −.49∗∗ −.44∗∗ −.44∗∗ −.52∗∗ — —
25. Satisfaction with life .36∗∗ .50∗∗ .34∗∗ .44∗∗ .58∗∗ .51∗∗ — —
26. Sleep efficiency −.14 .12 .27∗∗ .19∗∗ .06 .09 — —
27. Depressive — — — — — — — −.36∗∗
symptomatology (BSI
derived)
28. Anxiety (BSI derived) — — — — — — — −.19∗∗
29. Somaticism (BSI — — — — — — — −.16∗∗
derived)
30. Hostility (BSI derived) — — — — — — — −.20∗∗
Note. Sample 1 = community-based sample of younger adults (n = 193).
Sample 2 = female osteoarthritis patients (n = 183).
Sample 3 = male spouses of osteoarthritis patients (n = 168).
Sample 4 = community-based sample of middle-aged women (n = 378).
Sample 5 = women with early stage breast cancer (n = 198).
Sample 6 = women with late stage breast cancer (n = 86).
Sample 7 = undergraduate students (n = 359).
Sample 8 = undergraduate students (n = 511).
∗∗ Correlation is significant at the 0.01 level (2-tailed).
∗ Correlation is significant at the 0.05 level (2-tailed).

that is critical to defining the construct. To deter- to a new sample, to determine which purpose in
mine the central role played by this aspect of life life scale provided the better prediction of subjec-
purpose, we administered the LET and another com- tive well-being. In that the Purpose in Life Scale
monly used measure of purpose in life, the 9-item has some of the same limitations as other existing
version of Ryff’s (Ryff, 1989) Purpose in Life Scale purpose in life scales, e.g., the presence of items
296 Scheier et al.

Table V. Zero-Order Correlations and Partial Correlations Between the

Life Engagement Test, Ryff’s (Ryff, 1989) Purpose in Life Scale, and Indica-
tors of Subjective Well-Being
Zero-order
correlations Partial correlations

Subjective well-being LET PILS LET PILS

Life satisfaction .51∗∗ .44∗∗ .31∗∗ .11
Self-esteem .59∗∗ .49∗∗ .38∗∗ .11
Self-mastery .47∗∗ .46∗∗ .23∗∗ .18∗
Positive affect .59∗∗ .60∗∗ .27∗∗ .31∗∗
Negative affect −.28∗∗ −.16∗ −.24∗∗ .06
Depressive symptomatology −.42∗∗ −.37∗∗ −.23∗∗ −.10
Perceived stress −.46∗∗ −.33∗∗ −.33∗∗ .01
Note. LET = Life Engagement Test and PILS = Purpose in Life Scale. The
partial correlations between the Life Engagement Test and the indicators of
subjective well-being were controlled for the Purpose in Life Scale, and vice
versa.
∗∗ p <.01; ∗ p <.05.

measuring related constructs and the framing of DISCUSSION

items in a way that makes it hard to assess changes
in purpose in life over time, we expected that the The purpose of this brief report is to provide be-
LET would be a better predictor of well-being than havioral medicine and health psychology researchers
would the Purpose in Life Scale. The sample stud- with a new scale that can be used to assess purpose in
ied (separate from the eight samples described previ- life. The content of the scale is shaped by recent the-
ously) consisted of 137 undergraduates and the sub- ories of behavioral self-regulation, which suggest that
jective well-being indicators included life satisfaction purpose in life is largely derived from having valued
(Diener et al., 1985), self-esteem (Rosenberg, 1965), activities in which to engage. A large volume of data
self-mastery (Pearlin and Schooler, 1978), positive was presented examining the factor structure of the
and negative affect experienced during the past year scale, its internal consistency, test-retest reliability,
(Watson et al., 1988), depressive symptomatology and associations with health-relevant variables and
(Radloff, 1977), and perceived stress (Cohen et al., other psychosocial factors many of which have been
1983). shown to have important implications for health. The
The correlation between the LET and the Pur- LET was also shown to be more strongly related to
pose in Life Scale was .73. As Table V shows, sig- indicators of subjective well-being than is at least
nificant correlations emerged between the LET and one commonly used existing scale of purpose in life.
the Purpose in Life Scale and each of the sub- There was little evidence of gender, age, or ethnic dif-
jective well-being indicators assessed. More impor- ferences on the LET (although African Americans
tantly, the correlations between the LET and the did tend to score higher than Caucasians). Taken to-
subjective well-being indicators all remained signif- gether, the data assembled suggest that the LET is
icant when the analyses controlled for the Purpose sound psychometrically and is ready to be used in
in Life Scale. In contrast, when the LET was con- research.
trolled, only two of the seven associations between
the Purpose in Life Scale and the indicators of sub- Relevance to Behavioral Medicine
jective well-being remained significant. This pattern and Health Psychology
of results remained stable if simultaneously control-
ling for participants’ gender, age, and race. Taken to- Clearly, life engagement is a broad construct
gether, these data strongly suggest that it is the com- that could be of interest to a wide variety of psy-
ponent of purpose in life that is measured by the chosocial researchers. Why then publish the Life En-
LET that is important in determining the relation- gagement Test here? Although we believe that the
ships between purpose in life and subjective well- scale may be of interest to researchers more gener-
being. ally, we are most interested in seeing the scale used
The Life Engagement Test: Assessing Purpose in Life 297

by those doing work in health psychology and be- Carver, C. S., and Scheier, M. F. (1981). Attention and Self-
havioral medicine. In this paper, we have shown that regulation: A Control-Theory Approach to Human Behavior.
Springer Verlag, New York.
life engagement correlates with a variety of other Carver, C. S., and Scheier, M. F. (1990). Origins and functions of
psychosocial factors, such as dispositional optimism, positive and negative affect: A control-process view. Psychol.
social network size, and emotional expression style, Rev. 97: 19–35.
Carver, C. S., and Scheier, M. F. (1998). On the Self Regulation of
which are known to link to health outcomes. We Behavior. Cambridge University Press, New York.
have also shown that life engagement is associated Cohen, S., Doyle, W. J., Skoner, D. P., Rabin, B. S., and Gwaltney,
with self-rated health and health-related physical and J. M. (1997). Social ties and susceptibility to the common cold.
J. Am. Med. Assoc. 277: 1940–1944.
mental functioning. Cohen, S., Kamarck, T., and Mermelstein, R. (1983). A global
We are currently exploring associations between measure of perceived stress. J. Health. Soc. Behav. 24: 385–
life engagement and selected biological and disease 396.
Cohen, S., Mermelstein, R., Kamarck, T., and Hoberman, H.
outcomes, and some of these preliminary analyses (1985). Measuring the functional components of social sup-
are proving quite interesting. For example, in Sam- port. In Sarason, I. G., and Sarason, B. (Eds.), Social Sup-
ple 4, high LET scores are associated with lower lev- port: Theory, Research and Applications, Martinus Nijhoff,
The Hague, pp. 73–94.
els of aortic calcification, controlling for standard car- Cook, W. W., and Medley, D. M. (1954). Proposed hostility and
diovascular disease risk factors and many of the psy- pharisaic-virtue scales for the MMPI. J. Appl. Psychol. 38:
chosocial variables presented in Table IV for Sam- 414–418.
Derogatis, L. R., and Melisaratos, N. (1983). The brief symptom
ples 1 through 6 (Matthews et al., 2005). Although inventory: An introductory report. Psychol. Med. 13: 595–
admittedly only a beginning, we hope that these 605.
findings (when taken together) offer the enticement Diener, E., Emmons, R. A., Larson, R. J., and Griffin, S. (1985).
The satisfaction with life scale. J. Pers. Assess. 49: 71–75.
needed to convince other behavioral medicine and Feather, N. T. (Ed.). (1982). Expectations and Actions: Expect-
health psychology researchers to incorporate the Life ancy-Value Models in Psychology. Erlbaum, Hillsdale,
Engagement Test into their own protocols in order to NJ.
Goldberg, L. R. (1992). The development of markers for the big-
further explore associations between life engagement five factor structure. Psychol. Assess. 4: 26–42.
and physical well-being. Kaplan, R. M. (1990). Behavior as the central outcome in health
care. Am. Psychol. 45: 1211–1220.
Locke, H., and Wallace, K. (1959). Short marital-adjustment and
ACKNOWLEDGMENTS prediction tests: Their reliability and validity. Marriage Fam.
Living 21: 251–255.
Matthews, K. A., Owens, J. F., Edmunowicz, D., and Kuller,
Preparation of this article was supported by L. H. (2005). Positive and negative affect/cognitions and
funds awarded to the Pittsburgh Mind-Body Center risk for coronary and aortic calcification in healthy women.
at the University of Pittsburgh and Carnegie Mellon Manuscript under review.
Pearlin, L. I., and Schooler, C. (1978). The structure of coping.
University (NIH HL65111, HL65112, HL76858, and J. Health. Soc. Behav. 19: 2–21.
HL76852). Preparation of this article also benefited Radloff, L. S. (1977). The CES-D scale: A self-report depression
from the assistance offered by Suzanne Colvin and scale for research in the general population. Appl. Psychol.
Meas. 1: 385–401.
Ginger Placone, and the authors thank them for their Rosenberg, M. (1965). Society and Adolescent Self-Image. Prince-
help. ton University Press, Princeton, NJ.
Ryff, C. D. (1989). Happiness is everything, or is it? Explorations
on the meaning of psychological well-being. J. Pers. Soc. Psy-
chol. 57: 1069–1081.
REFERENCES Scheier, M. F., Carver, C. S., and Bridges, M. W. (1994).
Distinguishing optimism from neuroticism (and trait anxi-
ety, self-mastery, and self-esteem): A re-evaluation of the
Andresen, E. M., Malmgren, J. A., Carter, W. B., and Patrick, Life Orientation Test. J. Pers. Soc. Psychol. 67: 1063–
D. L. (1994). Screening for depression in well older adults: 1078.
Evaluation of a short form of the CES-D (Center for Epi- Shah, J., and Higgins, E. T. (1997). Expectancy X value effects:
demiologic Studies Depression Scale). Am. J. Prev. Med. 10: Regulatory focus as determinant of magnitude and direction.
77–84. J. Pers. Soc. Psychol. 73: 447–458.
Atkinson, J. W. (1964). An Introduction to Motivation. Van Nos- Shrout, P. E., and Yager, T. J. (1989). Reliability and validity of
trand, Princeton, NJ. screening scales: Effect of reducing scale length. J. Clin. Epi-
Barefoot, K. C., Dodge, K. A., Peterson, B. L., Dahlstrom, W. G., demiol. 42: 69–78.
and Williams, R. B. (1989). The Cook-Medley Hostility Scale: Spielberger, C. D., et al. (1985). The experience and ex-
Item content and ability to predict survival. Psychosom. Med. pression of anger: Construction and validation of
62: 17–25. an anger expression scale. In Chesney, M. A., and
Buysse, D. J., Reynolds, C. F., Monk, T. H., Berman, S. R., and Rosenman, R. H. (Eds.), Anger and Hostility in Car-
Kupfer, D. J. (1989). The Pittsburgh sleep quality index. Psy- diovascular and Behavioral Disorders. Hemisphere,
chiatry. Res. 28: 193–213. Cambridge.
298 Scheier et al.

Vroom, V. H. (1964). Work and Motivation. Wiley, New York. Wrosch, C., Scheier, M. F., Carver, C. S., and Schulz, R. (2003a).
Ware, J. E., Sherbourne, C. D. (1992). The MOS 36- The importance of goal disengagement in adaptive self-
Item Short-Form Health Survey (SF-36): I. Concep- regulation: When giving up is beneficial. Self Identity 2: 1–20.
tual framework and item selection. Med. Care 30: 473– Wrosch, C., Scheier, M. F., Miller, G. E., Schulz, R., and
483. Carver, C. S. (2003b). Adaptive self-regulation of unattain-
Watson, D., Clark, L. A., and Tellegen, A. (1988). Development able goals: Goal disengagement, goal re-engagement, and
and validation of brief measures of positive and negative af- subjective well-being. Per. Soc. Psychol. Bull. 29: 1494–
fect: The PANAS scales. J. Pers. Soc. Psychol. 54: 1063–1070. 1508.