Copyright 1994 by the American Psychological Association, Inc.

Journal of Personality and Social Psychology

1994, Vol. 67, No. 6, 1063-1078 0022-3514/94/S3.00

Distinguishing Optimism From Neuroticism (and Trait Anxiety,

Self-Mastery, and Self-Esteem): A Reevaluation of the
Life Orientation Test
Michael F. Scheier, Charles S. Carver, and Michael W. Bridges

Research on dispositional optimism as assessed by the Life Orientation Test (Scheier & Carver, 1985)
has been challenged on the grounds that effects attributed to optimism are indistinguishable from
those of unmeasured third variables, most notably, neuroticism. Data from 4,309 subjects show that
associations between optimism and both depression and aspects of coping remain significant even
when the effects of neuroticism, as well as the effects of trait anxiety, self-mastery, and self-esteem,
are statistically controlled. Thus, the Life Orientation Test does appear to possess adequate predictive
and discriminant validity. Examination of the scale on somewhat different grounds, however, does
suggest that future applications can benefit from its revision. Thus, we also describe a minor modi-
fication to the Life Orientation Test, along with data bearing on the revised scale's psychometric
properties.

Accumulating evidence from a variety of sources suggests tendencies (Carver, Scheier, & Weintraub, 1989) and in the
that dispositional optimism is beneficial for physical and psy- kinds of coping responses that they spontaneously generate
chological well-being. For example, Aspinwall and Taylor when given hypothetical coping situations (Scheier, Weintraub,
(1992) have recently shown that optimistic persons adjust more & Carver, 1986). Optimists also differ from pessimists in the
favorably to important life transitions than do persons who are manner in which they cope with serious disease (Friedman et
more pessimistic in outlook. In a similar vein, Litt, Tennen, al., 1992) and with concerns about specific health threats (e.g.,
Affleck, and Klock (1992) have reported that optimistic women Carver etal., 1993; Stanton& Snider, 1993; Taylor etal., 1992).
who are unsuccessful at in vitro fertilization respond better to A general characterization of thefindingsof this research is that
the failure than do women who are more pessimistic. Concep- optimists tend to use more problem-focused coping strategies
tually similar results have also been reported by Scheier et al. than do pessimists. When problem-focused coping is not a pos-
(1989). Their study tracked a group of men undergoing coro- sibility, optimists turn to more adaptive emotion-focused cop-
nary artery bypass surgery. Optimistic men evidenced a more ing strategies such as acceptance, use of humor, and positive
rapid physical recovery after their surgery and reported a higher reframing of the situation. Pessimists tend to cope through overt
quality of life 6 months postoperatively than did the more pes- denial and by mentally and behaviorally disengaging from the
simistic men in the sample. Nor are these the only beneficial goals with which the stressor is interfering, regardless of
effects for dispositional optimism that have been reported in whether something can be done to solve the problem or not.
the literature (for a more comprehensive review, see Scheier &
These findings regarding optimism are consistent with the
Carver, 1992).
model of behavioral self-regulation from which our own work
Related research suggests that these differences in outcomes on optimism grew (e.g., Carver & Scheier, 1981, 1990a; Scheier
derive partly from differences between optimists and pessimists & Carver, 1988). This is a model that has roots in the long tra-
in the manner in which they cope with the challenges in their dition of expectancy-value theories in psychology. In this
lives. Optimists differ from pessimists in their stable coping model, people are seen as remaining engaged in efforts to over-
come adversity to reach goals as long as their expectancies of
eventual success are sufficiently favorable. When their doubts
Michael F. Scheier and Michael W. Bridges, Department of Psychol-
are too severe, people are more likely to give up on the threat-
ogy, Carnegie Mellon University; Charles S. Carver, Department of Psy-
chology, University of Miami. ened goals. These differences in expectancies are also assumed
Data collection for this article was supported by National Science to be paralleled by variations in affective experience (for details,
Foundation (NSF) Grants BNS87-17783 and BNS87-06271. Prepara- see Carver & Scheier, 1990b). With enough movement toward
tion of the article was facilitated by NSF Grants BNS90-11653 and desired goals, affect is positive. If movement toward desired
BNS90-10425, American Cancer Society Grant PBR-56, and National goals is sufficiently impeded, affect is negative.
Heart, Lung, and Blood Institute grant HL44436-01A1. Although this viewpoint on behavior and affect can be ap-
We are indebted to Joel Greenhouse for consulting on those aspects
plied in terms of situational variations in expectancies across
of the data analysis pertaining to factor analysis, and to Jeffrey Thomas
for assisting in the preparation of data tables. time or events, it can also be applied in terms of individual
Correspondence concerning this article should be addressed to Mi- differences. Optimists are people who tend to hold positive ex-
chael F. Scheier, Department of Psychology, Carnegie Mellon Univer- pectancies for their future; pessimists are people who tend to
sity, Pittsburgh, Pennsylvania 15213. hold more negative expectations for the future. Thus, our anal-
1063
 1064 M. SCHEIER, C. CARVER, AND M. BRIDGES

ysis of how optimism versus pessimism leads to different re- fied and measured reliably (Carver et al., 1989). The possibility
sponses to adversity is one application of a more general model thus remains that dispositional optimism may be uniquely re-
of the processes that underlie behavior, a model that is applica- lated in important ways to outcome variables that went un-
ble to a wide range of motivational issues and contexts. measured in these earlier studies. In this regard, it is interesting
to note that dispositional optimism was associated with eight
Critique and Challenge different coping qualities in Carver et al.'s (1993) study of ad-
justment to breast cancer surgery. Of these eight coping re-
Much of the research on optimism and pessimism (although sponses, only half were measured in the Smith et al. (1989)
certainly not all of it; e.g., Beck, Steer, Kovacs, & Garrison, study.
1985; Reker & Wong, 1983) has made use of the Life Orienta- Our third observation has to do with the nature of the pre-
tion Test (LOT; Scheier & Carver, 1985) to assess individual dictors with which optimism has been compared. Most of the
differences on this dimension. This scale has recently been crit- concern to date has centered around the overlap between opti-
icized by others. By implication, this criticism also undermines mism and neuroticism or negative affectivity, as indexed by one
the integrity of the optimism construct. The primary purpose or another measure of chronic anxiety. Note, however, that neu-
of this article is to address these issues. roticism is conventionally viewed as a multifaceted construct
Most of the criticism aimed at the scale involves the third that consists partly of the absence of optimism (i.e., pessimism).
variable problem. Smith, Pope, Rhodewalt, and Poulton (1989) Thus, there is a distinct conceptual link between constructs. On
were the first to raise this issue, doing so with respect to trait the other hand, neuroticism also incorporates a host of other
anxiety. That is, they questioned whether effects attributable to factors, such as self-doubt, emotional lability, and worry. Com-
optimism might really be due to variance that optimism shared bining qualities in this way can create problems of interpreta-
with trait anxiety. Consistent with this view, Smith et al. (1989) tion (Carver, 1989) because it becomes very difficult to identify
reported relatively high correlations between optimism and which components of neuroticism underlie a given effect. As a
trait anxiety across two independent samples. Smith et al. also hypothetical example, it might be the pessimism facet of neu-
showed that it was possible to eliminate the significant negative roticism that relates to such variables as active coping, plan-
association that they found between optimism and reports of ning, giving up, and positive reinterpretation. The emotional
physical symptoms by controlling for the effects of trait anxiety. lability component may not be as good a predictor of these vari-
In contrast, the association between trait anxiety and reports of ables, but may relate well to other variables such as the experi-
physical symptoms remained significant even after the effects ence of physical symptoms.
of dispositional optimism were removed. Similarly, significant The other constructs under consideration here also have con-
zero-order correlations between optimism and different varie- ceptual overlap with optimism, though in a different way. Self-
ties of coping were largely eliminated when the effects of trait mastery is the perception that one exerts control over the events
anxiety were controlled, whereas many of the associations be- in one's life (Pearlin & Schooler, 1978). This construct thus in-
tween anxiety and coping remained strong after removing the corporates a strong sense of positive expectancy for the future,
effects of dispositional optimism. In a similar vein, Marshall but weds to it a sense of personal responsibility for that expec-
and Lang (1990) have also raised the third variable problem, tancy. Self-esteem shares ground with optimism in a more
but with respect to self-mastery rather than trait anxiety (see diffuse way. Self-esteem represents a sense of self-worth, which
also Robbins, Spence, & Clark, 1991). carries the implication that one will be accepted rather than
We have several observations to make with respect to this rejected by others, and that one is not a failure in one's life.
work. Ourfirstpoint concerns the nature of the outcome vari- These consequences, of course, involve positive versus negative
ables that have been examined across studies. Both Smith et outcomes, thus linking self-esteem conceptually to optimism.
al. (1989) and Robbins et al. (1991) examined the relationship As with self-mastery, what seems to differentiate this concept
between optimism and health complaints, and neither found an from optimism involves (at least in part) a kind of ascription to
independent effect for optimism when variables such as trait the self. The ascription in this case, however, is not one of con-
anxiety were controlled. Smith et al. also examined coping trol but rather of an intrinsic tie between feelings of worth or
strategies, however, and found that optimism was an indepen- the self's value and positive outcomes.
dent predictor of certain coping responses. Similarly, Robbins In sum, the alternative constructs being examined here all
et al. (1991) also examined health maintenance behaviors and have conceptual as well as empirical overlap with optimism.
found optimism to be an independent predictor of these. The Each, however, incorporates at least one additional quality that
point here is that optimism may be a stronger independent pre- takes it beyond optimism. In the case of neuroticism, there may
dictor of some outcomes than of others. Shared variance may be several such qualities.
explain the association between optimism and symptom report- Our fourth andfinalobservation is that results pertaining to
ing, for example, but may not fully explain the link between the discriminant validity of the LOT have not all been com-
optimism and other outcomes of interest. pletely negative. As noted earlier, Smith et al. (1989) found that
There is a second (though related) issue here, as well. Previous the associations between self-blame and optimism (in both of
researchers have used outcome measures that were somewhat their studies) and seeking of social support and optimism (in
limited in scope. For example, Smith et al. (1989) relied on only one of their studies) remained significant even when trait anxi-
five coping categories to explore associations between disposi- ety was controlled. Robbins et al. (1991) found that the associ-
tional optimism and coping tendencies, controlling for trait anx- ation between optimism and health maintenance behaviors re-
iety. Yet a far greater number of coping responses can be identi- mained intact, even when the effects of manifest anxiety, instru-
 REEVALUATING THE LIFE ORIENTATION TEST 1065

mentality, anger, and alienation were simultaneously controlled. agreement with statements such as "In uncertain times, I usually expect
the best," and "I hardly ever expect things to go my way," using a 5-
Indeed, of all the variables studied, optimism was one of only
point response scale ranging from 0 (strongly disagree) to 4 (strongly
two that made significant independent contributions to the pre- agree). Of the 8 scored items, 4 are worded in a positive direction and 4
diction of health maintenance behaviors. More recently, Aspin- are worded in a negative direction. After reversing the scoring for the
wall and Taylor (1992) have reported that optimism predicts negatively worded items, item scores were totaled to yield an overall
adjustment to the first semester of college, independent of self- optimism score with high scores representing greater optimism. In our
esteem, locus of control, and desire for control. Finally, Mroc- sample, scores ranged from 0 to 32. Cronbach's alpha was .82.
zek, Spiro, Aldwin, Ozer, and Bosse (1993) have found that op- Neuroticism. The Emotional Stability subscale of the Guilford-
timism continues to predict psychological distress among mid- Zimmerman Temperament Survey (GZTS; Guilford, Zimmerman, &
dle-aged men, even after the distress measure is adjusted for Guilford, 1976) was used to assess neuroticism. Participants were asked
differences in neuroticism. Given these various considerations, to indicate if a series of 30 statements were true for them, using a 3-
it seems premature to conclude that the LOT has no predictive point response scale (1 = yes, 2 = no,i = uncertain). Two sample items
validity independent of other measures. In the same way, it are "You are sometimes bubbling over with energy and sometimes very
seems premature to close the book on optimism-pessimism as sluggish," and "Your mood often changes from happiness to sadness, or
vice versa, without your knowing why." Item responses are first recoded
an independent construct.
as needed so that higher scores indicate higher levels of neuroticism. To
In this article we address this set of questions further. For the recode, responses receiving a score of 1 or 2 are reversed, responses
past several years, we have been collecting information from receiving a score of 3 are left the same. An overall neuroticism score is
large groups of respondents on a variety of personality variables, then computed by totaling the number of responses receiving a score of
coping styles, and other outcome measures. The personality bat- 1 (responses receiving a 2 or 3 do not contribute to the overall score).
tery includes a measure of optimism-pessimism (the LOT) as The Emotional Stability subscale of the GZTS has been recommended
well as measures of self-esteem, trait anxiety, self-mastery, and as a good measure of neuroticism by others (e.g., Costa & McCrae,
neuroticism. The measure of coping is broad in scope and cov- 1985). Cronbach's alpha for the present sample was .85.
ers a wide range of diverse coping tendencies. The other outcome Self-mastery. Self-mastery was assessed by using Pearlin and
measures include both physical symptoms (number and inten- Schooler's (1978) Self-Mastery Scale (SMS). This seven-item instru-
sity) and a measure of depression. ment assesses the extent to which a person generally feels as though
The primary purpose of this article is to use the data we have he or she manifests personal mastery over life outcomes (e.g., "What
happens to me in the future mostly depends on me," and "Sometimes
assembled to reexamine the predictive validity of the LOT by
I feel that I am being pushed around in life"). Its basic psychometric
using it and the other personality factors to predict variations in properties are well established (Pearlin & Schooler, 1978), and it has
coping, symptoms, and depression. Two general sets of predic- been used successfully in the past with several different community-
tions are advanced. First, on the basis of findings reported by based populations (e.g., Pearlin & Schooler, 1978). It was also the scale
Smith et al. (1989), Robbins et al. (1991), and Mroczek et al. that Marshall and Lang (1990) used in their study exploring the predic-
(1993), we expect that zero-order correlations between opti- tive power of optimism and self-mastery with respect to depression.
mism and physical symptoms will be substantially reduced Cronbach's alpha for the SMS in the present sample was .75.
when the data are adjusted for trait anxiety and neuroticism. Self-esteem. Rosenberg's (1965)10-item Self-Esteem Scale (or SES)
Second, we expect that other associations involving optimism, was used to assess self-esteem. The scale, which provides a convenient
coping, and depression will remain strong, even after the data measure of global attitudes about the self, has five negatively worded
are adjusted for the various personality factors that have been items and five positively worded items. Participants were asked to indi-
measured. cate their agreement on a scale of 1 (strongly disagree) to 4 (strongly
agree) with statements such as "I feel I have a number of good qualities,"
and "At times, I think I am no good at all." This scale is one of the most
Study 1: Reevaluating the Life Orientation Test widely used measures of self-esteem and has displayed good reliability
and validity (Crandall, 1973; Rosenberg, 1965). In our sample, the scale
Method had an internal reliability of .88.
Trait anxiety. The trait form of the State-Trait Anxiety Inventory
Subjects and Procedure
(STAI; Spielberger, Gorsuch, & Lushene, 1974) was used to measure
A total of 4,309 undergraduates from Carnegie Mellon University and trait anxiety. This scale is composed of 20 Likert items evaluating the
the University of Miami participated in the research (1,846 women, extent to which the respondents experience a variety of feelings such
2,417 men, and 46 participants who did not indicate their gender). Par- as happiness, self-confidence, tension, and disappointment (e.g., "I feel
ticipation was in partial fulfillment of a psychology research require- content," and "I worry too much over something that really doesn't
ment. AH subjects completed a number of scales as part of a larger pre- matter"). The scale has been used extensively in prior psychosocial re-
testing packet. Packets were administered in large group testing sessions search and its psychometric properties have been well documented (see,
across successive semesters from 1988 to 1990. Because of time con- e.g., Watson & Clark, 1984). Cronbach's alpha for the current sample
straints, not all groups received all scales. As a result, sample sizes for was .89.
analyses vary from analysis to analysis. Depression. The Beck Depression Inventory (BDI) short form
(Beck, Rial, & Rickels, 1974) was used to assess depression. The BDI
assesses attitudes and symptoms derived from clinical observations that
Measures
are typically observed in depressed psychiatric patients but not in non-
Optimism. Optimism was measured by using the LOT (Scheier & depressed psychiatric patients (Beck, Ward, Mendelson, Mock, & Er-
Carver, 1985). The LOT is an eight-item self-report measure (plus four baugh, 1961). The scale has been widely used and has well-established
filler items) assessing generalized expectancies for positive versus nega- psychometric properties (Beck, Steer, & Garbin, 1988). The short form,
tive outcomes. Respondents were asked to indicate their degree of 13-item version used in our study assessed attitudes and symptoms
 1066 M. SCHEIER, C. CARVER, AND M. BRIDGES

across several different domains, including (but not limited to) mood,
Table 1
sense of failure, lack of satisfaction, social withdrawal, and indecisive- Correlations Among Predictor Variables
ness. For each item, respondents were asked to choose from a group of
four statements (rated 0 to 3 in depressive symptomatology) the state- Variable 1 2 3 4 5
ment that best described the way they were feeling that day. Scores for
the present sample ranged from 0 to 39 (with higher scores indicating 1. Optimism
greater depression). Cronbach's alpha was .87. r — .55 -.59 -.50 .54
Physical symptoms. Self-reports of physical symptoms were as- N 1,883 1,420 1,692 595
2. Self-mastery
sessed with the Cohen-Hoberman Inventory of Physical Symptoms
r — -.69 -.48 .58
(CHIPS; Cohen & Hoberman, 1983). The CHIPS comprises a list of 38
N 572 569 624
commonly experienced physical symptoms (e.g., back pain, headache, 3. Trait anxiety
and stuffy head or nose). Symptoms more psychological in nature (e.g., r — .74 -.72
feeling depressed or anxious) are explicitly excluded from the list. Sub- N 1,423 181
jects were asked to indicate how much they had been bothered by each 4. Neuroticism
of the symptoms in the past 4 weeks, using a 5-point scale (1 = not at all r -.66
to 5 = extremely). In past research (Cohen & Hoberman, 1983), the N 181
CHIPS was found to predict use of student health facilities among two 5. Self-esteem
separate samples of college students (Cohen & Hoberman, 1983), r
N
thereby supporting the construct validity of the scale.
Responses to the CHIPS were scored in two ways in our research. Note. All correlations specified here reached significance at p < .001,
First, a measure of symptom intensity was computed by summing the two-tailed.
degree to which subjects reported being bothered by each symptom
across all 38 symptoms. Scores for symptom intensity ranged from 38
to 162. Second, a simple tally was made of the number of symptoms for
which subjects indicated they were bothered to some degree, ignoring Is Optimism a Predictor of Outcomes?
the extent to which they were bothered (i.e., a score of 1 was given for
each item that received a response of 2 or more). Scores for number of Zero-order correlations between predictors and outcomes are
symptoms ranged from 0 to 35 in the present sample. shown in Table 2. As can be seen, all of the predictors were
Coping. Coping was measured by the COPE (Carver et al., 1989). moderately correlated with depression to about the same de-
The COPE is a 60-item, multidimension coping instrument designed to gree. Depression correlated highest with self-esteem and lowest
assess 15 conceptually distinct methods of coping. The 60 items repre-
with neuroticism. In addition, all of the predictors were signifi-
sent a large range of coping responses including (but not limited to)
active coping, positive reinterpretation and growth, seeking of social
cantly correlated with the two symptom measures. For symp-
support for emotional reasons, denial, mental and behavioral disengage- tom intensity, the highest correlations involved neuroticism
ment, and focusing on and venting of emotions (e.g., "I do what has to and trait anxiety. The lowest correlation involved optimism. A
be done, one step at a time," "I turn to work or other substitute activities similar ordering occurred for number of symptoms.
to take my mind off things," and "I talk to someone about how I feel"). As a group, the predictor variables were also substantially
Participants were instructed to indicate how much they usually did each correlated with different aspects of coping. All of the predictors
of the things that the items reflected when they encountered difficulties correlated significantly with active coping, planning, positive re-
or problems in their lives, using a 4-point Likert scale ( 1 = 7 usually interpretation and growth, denial, mental disengagement, and
don't do this at all to 4 = I usually do this a lot). Cronbach's alpha for behavioral disengagement. Correlations between predictors and
the 15 scales in the current study ranged from .37 (mental disengage-
outcomes tended to be higher for the more negative coping ten-
ment) to .93 (use of religion). With the exception of mental disengage-
ment, the remainder of the alphas were all above .59, with the majority
dencies (e.g., mental and behavioral disengagement) than for
of the scales above .70. The average alpha across the 15 scales was .73. the more positive coping tendencies (e.g., active coping and
planning).
With the exception of dispositional optimism, associations
between the predictors and the remaining coping tendencies
Results
were more sporadic. For example, whereas four of thefivepre-
Correlations Among Predictors
1
Portions of the analyses from Study 1 have been briefly described
The zero-order correlations among the five predictor vari- elsewhere (in Scheier et al., 1989, and in Scheier and Carver, 1992).
ables are shown in Table I.1 As can be seen, all of the intercor- 2
Subsidiary analyses conducted separately by gender suggested that
relations among the predictors were significant. It is also note- the associations between the predictors tended to be somewhat higher
worthy, however, that the magnitudes of the correlations be- for women than for men, often significantly so, given the size of the
tween the LOT and the other predictors were only in the samples. This pattern corresponds to the gender differences reported by
Scheier and Carver (1985) on a similar set of data. Subsidiary analyses
moderate range (the average correlation between the LOT and
were always conducted on the data to assess the effects of gender. Very
the other predictors was .54, which was the lowest average cor- few other gender differences emerged, however, in either Study 1 or
relation that was obtained). This generally suggests that the Study 2. For example, of the 72 partial correlations reported in Table 2,
LOT had less in common with the other predictors than the only one involved a significant difference due to gender. The few gender
other predictors had in common with each other, with the pos- differences that did emerge seemed random in nature and were basically
sible exception of self-mastery, which also had a relatively low uninterpretable. As a result, gender is discussed only when there are
average correlation (.56).2 meaningful differences to report.
 REEVALUATING THE LIFE ORIENTATION TEST 1067

Table 2
Correlations Between Predictor Variables and Outcomes

Outcomes Optimism Self-mastery Trait anxiety Self-esteem Neuroticism

Depression
r -.42*** -.43*** -.49*** -.54*** .41***
N 1,900 1,306 547 604 545
Number of symptoms
r -.21*** -.27*** .47*** -.26*** .51***
N 1,015 443 315 443 591
Intensity of symptoms
r -.25*** -.28*** 47 *»* -.27*** .54***
N 1,015 443 315 443 591
Active coping
r .30*** .32*** -.28*** .25*** -.20***
N 813 375 394 375 393
Planning
r .30*** 27*** -.17*** .20*** -.10*
N 813 375 394 375 393
Suppression of competing
activities
r .14*** .09 .00 .07 .02
N 815 375 396 375 395
Restraint
r .12*** .08 -.19*** .19* -.17*
N 814 375 395 375 394
Positive reinterpretation
and growth
r .47*** .34*** -.23*** .33*** -.20***
N 815 375 395 375 394
Use of humor
r .10** .10 -.13** .12* -.10
N 815 375 396 375 395
Seeking instrumental social
support
r .16*** .10 -.02 .08 .05
N 814 375 396 375 395
Seeking emotional social
support
r .12*** .11* .02 .07 .11*
N 815 375 396 375 395
Turning to religion
r .22*** .08 -.03 .06 -.05
N 816 375 397 375 396
Acceptance
r .10** .11* -.15** .04 -.10*
N 816 375 397 375 396
Denial
r -.17*** -.17*** .32*** -.20*** .26***
N 815 375 397 375 396
Focusing on and venting of
emotions
r -.10** -.09 .32*** -.10 .42***
N 815 375 396 375 395
Mental disengagement
r -.18*** -.23*** .34*** -.17*** .42***
N 816 375 397 375 396
Behavioral disengagement
r .33*** -.41*** .45*** -.38*** 37*.*
N 816 375 397 375 396
Use of drugs or alcohol
r -.11** .16** .27*** -.08 .19***
N 816 375 397 375 396

*p< .05, two-tailed. **p< .01, two-tailed. ***p< .001, two-tailed.

1068 M. SCHEIER, C. CARVER, AND M. BRIDGES

dictors (self-mastery, trait anxiety, neuroticism, and optimism) under consideration. In the case of partial correlations, the sole
were all significantly associated with drinking and substance concern is whether or not optimism as assessed by the LOT is a
abuse, only optimism correlated significantly with suppression unique and independent predictor of outcomes. There is an-
of competing activities, seeking instrumental social support, other way to address the contribution being made by optimism,
and turning to religion. Indeed, of thefivepredictors, only opti- however, and that is through the use of factor analysis. That is,
mism related significantly to every coping tendency that was recall our suggestion that characteristics such as neuroticism
assessed. Undoubtedly, this was due in part to the greater num- are multifaceted, and that optimism comprises only one of
ber of subjects who completed the LOT, but it was also due in many different facets. The critical question, then, is to deter-
part to the greater magnitude of the correlations. mine the importance of the optimism-pessimism component
in producing the associations that are observed. One way of get-
Is Optimism a Unique Predictor of Outcomes? ting information relevant to this point is to factor analyze the
items on the LOT along with the items contained on the other
Thefirstfour columns of Table 3 present partial correlations predictors. If the optimism component is important, the opti-
between optimism and the outcome variables, controlling for mism items ought to emerge as distinct, factor analytically, from
each of the other predictors in turn. As can be seen, many of the the items comprising the other scales. Moreover, factor scores
zero-order correlations involving optimism remain significant computed on the basis of such an optimism factor should ac-
even after the other predictors are statistically controlled, albeit count for a significant amount of variance in the outcomes.
the correlations are usually reduced in magnitude. This charac- To explore these possibilities, several sets of factor analyses
terization holds for associations involving depression and many were conducted on the data set. One set of analyses (conducted
aspects of coping, including active coping, planning, suppres- on a sample of 1,312) grouped together items from the LOT,
sion of competing activities, positive reinterpretation and GZTS, and STAI. The second set of analyses (conducted on a
growth, seeking social support for instrumental reasons, and sample of 845) grouped together items from the LOT, SES, and
turning to religion. Taken in isolation, none of the other predic- SMS. Separate analyses were run on these two subsets of items
tors was able to render the correlations between optimism and because an insufficient number of subjects had been adminis-
these various outcomes nonsignificant. tered allfivepredictors. Thus, a combined factor analysis would
Associations involving number of symptoms and symptom not have been meaningful. The predictors were grouped as they
intensity fared less well, particularly the associations involving were because self-esteem and self-mastery tend to measure
symptom number. That is, with respect to symptom number, more positive aspects of behavior, whereas trait anxiety and
the correlation with optimism was reduced to nonsignificance emotional instability tend to measure aspects of behavior that
if any of the other predictors was statistically controlled. With are more negative in nature (cf. Marshall et al., 1992).
respect to symptom intensity, inclusion of either trait anxiety or Initial analyses were conducted, using a principal compo-
neuroticism reduced the correlation with optimism to near nents extraction technique, followed by a Varimax rotation to
zero. It is interesting that the correlation between optimism and achieve afinalsolution. The number of factors retained for final
symptom intensity remained significant even when the data rotation in these analyses was determined by setting the eigen-
werefirstadjusted individually for self-mastery and self-esteem. value at 1.0. The item set containing the LOT, GZTS, and STAI
The last column of Table 3 presents partial correlations be- yielded 13 factors. Five factors emerged from the item set con-
tween optimism and the outcomes, controlling for all four of taining the LOT, SES, and SMS. Thefindingsfrom the two data
the predictors simultaneously. This obviously is a much more sets were highly similar, however, with respect to the items com-
stringent test of the unique predictive power of dispositional op- posing the LOT. In each case, two factors emerged, the first of
timism, for the simple reason that as covariates are added there which was defined by the positively worded LOT items (i.e., the
is less and less residual variance to predict. As can be seen, how- optimism items), the second of which was composed of the neg-
ever, significant partial correlations were obtained between op- atively worded LOT items (i.e., the pessimism items).34 In each
timism and planning, positive reinterpretation and growth, case, the absolute loading for each LOT item was above .58 on
seeking support for instrumental reasons, and turning to reli- its relevant factor (the average loading for the LOT items across
gion, even when the associations were adjusted for all four of the the two factors and across the two data sets was .69).
other predictors at once. Associations between optimism and Although these initial analyses yielded interpretable results
active coping and between optimism and suppression of com- with respect to the items from the LOT, they were not totally
peting activities remained close to significance as well. We find satisfactory for other reasons. Most important, there seemed to
these results particularly striking and quite inconsistent with be an over-extraction of factors. That is, for each data set, factors
the notion that dispositional optimism as measured by the LOT emerged that were defined by only one or two items each. As a
is nothing more than trait anxiety, self-esteem, self-mastery, or result, follow-up analyses were performed, in which only three
neuroticism under a different label. factors were retained for final rotation. The decision to retain

Does Optimism Emerge as a Distinct Factor Among the 3

It is not uncommon for the positively and negatively worded Life
Items in This Item Set, and Is This Factor a Predictor of Orientation Test items to load on separate factors. We have obtained
Outcomes? such solutions in the past (Scheier & Carver, 1985), as have others (e.g.,
Marshall, Wortman, Kusulas, Hervig, & Vickers, 1992).
4
The partial correlations just reported begin to address the An item was said to define a factor if it loaded higher on that factor
issue of the independence of optimism from the other concepts than on any other factor in the analysis.
 REEVALUATING THE LIFE ORIENTATION TEST 1069

Table 3
Partial Correlations Between Optimism and Outcomes
Variable controlled for

Outcome Self-mastery Trait anxiety Self-esteem Neuroticism All variables

Depression
r -.24*** -.19*** -.18*** .28*** -.11
N 1,303 544 592 542 175
Number of symptoms
r -.08 .09 -.09 .06 .04
N 440 312 312 588 175
Intensity of symptoms
r -.12** .04 -.12** .03 .09
N 440 312 440 588 175
Active coping
r .16** .17*** .20*** 24*** .13
N 372 391 372 390 175
Planning
r .19*** .25*** .23*** .29*** .20**
N 372 391 372 390 175
Suppression of competing
activities
r .11* .17*** .13* .18*** .14
N 372 393 372 392 175
Restraint
r .09 .01 .07 .07 .03
N 372 392 372 391 175
Positive reinterpretation
and growth
r .36*** .43*** .37*** .43** .36***
N 372 392 372 391 175
Use of humor
r .06 .03 .04 .06 .02
N 372 393 372 392 175
Seeking instrumental social
support
r .13* .19*** .14** .21*** .16*
N 372 393 372 392 175
Seeking emotional social
support
r .08 .17*** .11* .21*** .13
N 372 393 372 392 175
Turning to religion
r .22*** .25*** .23*** .23*** .24***
N 372 394 372 393 175
Acceptance
r .05 .01 .10 .06 .03
N 372 394 372 393 175
Denial
r -.10 .02 -.08 -.06 .01
N 372 394 372 393 175
Focusing on and venting of
emotions
r -.06 .12* -.06 .14** .09
N 372 393 372 392 175
Mental disengagement
r -.07 .02 -.11* .03 .03
N 372 394 372 393 175
Behavioral disengagement
r -.14** -.10 -.17*** -.19*** -.04
N 372 395 372 393 175
Use of drugs and alcohol
r -.03 .07 -.08 -.02 .04
N 372 394 372 393 175

*p< .05, two-tailed. **p< .01, two-tailed. ***p< .001, two-tailed.

1070 M. SCHEIER, C. CARVER, AND M. BRIDGES

three factors was prompted by two considerations. First, the knowing why," and "You have frequent ups and downs in mood,
items on which each data set was based came from three sepa- sometimes with and sometimes without apparent cause." This
rate scales. Second, scree tests (Cattell, 1966) on the analyses Emotional Instability factor accounted for an additional 5.2%
just reported suggested the viability of a three-factor solution in of the variance. The final factor to emerge from this analysis
each case. involved feelings of worry, dread, and excessive rumination. It
The results of these three-factor analyses yielded simpler and was defined by eight items from the GZTS and seven items from
more interpretable solutions. With respect to the analysis in- the STAI. Illustrative items loading high on the Worry factor
volving the LOT, SES, and SMS, the first factor was denned by include "Disappointments affect you so little that you seldom
nine items from the SES and one item from the SMS (which think about them twice" (negatively weighted), "You give little
loaded lowest of the nine items). This Self-esteem factor ac- thought to your failures after they are just passed" (negatively
counted for 31.9% of the variance. The second factor was totally weighted), and "I worry too much over something that doesn't
denned by the eight items from the LOT. Factor loadings for matter." The Worry factor accounted for an additional 3.5% of
the LOT items on this Optimism factor ranged from .49 to .74 the variance among the items in this item set.
(averaging .60). The Optimism factor accounted for an addi- To determine whether the Optimism factors in these data sets
tional 8.1% of the variance. Thefinalfactor was denned by six contributed significantly to the prediction of outcomes, we gen-
items from the SMS and one item from the SES (which loaded erated factor scores reflecting the Optimism factor for each set
second lowest of the seven items). This Self-mastery factor ac- of data. Because the Varimax final rotations were based on ini-
counted for an additional 6.2% of variance. tial principal components analyses, the factor scores defining
In the analysis involving the LOT, GZTS, and STAI, the first the Optimism factor in each data set should have been statisti-
factor was defined by 23 items (8 items from the LOT, 9 items cally independent from the factor scores defining the remaining
from the STAI, and 6 items from the GZTS). GZTS and STAI two factors. Consistent with this expectation, intercorrelations
items loading high on this factor tended either to reflect explic- among the three sets of factor scores from each set of data were
itly an optimistic orientation (e.g., "You have usually been opti- all found to be zero. Thefirstcolumn of Table 4 presents corre-
mistic about your future") or reflect a positive mood state (e.g., lations between Optimism factor scores and outcomes for the
"I am happy"). The eight LOT items were among the 11 highest data set involving the LOT, SMS, and SES (correlations between
loading items on this factor (factor loadings for the LOT items outcomes and scores on the Self-esteem and Self-mastery fac-
ranged from .51 to .71, averaging .61). The three highest loading tors are also presented in Table 4). As can be seen, the Opti-
items on this factor all came from the LOT. This Optimism fac- mism factor scores fared fairly well in this data set, correlating
tor accounted for 23.1% of the variance. The second factor was significantly with 12 of the 18 outcomes that were assessed. Ta-
defined by 20 items (16 from the GZTS and 4 from the STAI). ble 5 presents correlations between Optimism factor scores and
Given the high number of GZTS items defining the factor, it is outcomes for the data set involving the LOT, STAI, and GZTS
not surprising that the factor tended to reflect emotional labil- (correlations between outcomes and scores on the Emotional
ity. The two highest loading items were "Your mood often Lability and Worry factors are also presented in Table 5). Opti-
changes from happiness to sadness, or vice versa, without your mism factor scores fared even better in this data set, correlating

Table 4
Correlations Between (Positive Scale) Factor Scores and Outcomes
Factor

Outcome Optimism Self-Esteem Self-Mastery

Depression -.27*** -.38*** -.38***

Intensity of symptoms -.05 -.16** -.24***
Number of symptoms -.09 -.14*** -.25***
Active coping .27*** .16** .22***
Planning .29*** .12* .15**
Suppression of competing activities .19*** .02 .00
Restraint coping .12* .09 .02
Positive reinterpretation and growth .47*** .18*** .17***
Use of humor .13* .08 .00
Seeking instrumental social support .18*** .05 .00
Seeking emotional social support .12* .06 .01
Turning to religion .29*** -.03 .03
Acceptance .10 .00 .07
Denial -.05 -.17*** -.18***
Focusing on and venting of emotions -.07 -.03 -.11*
Mental disengagement -.15** -.06 -.23***
Behavioral disengagement -.20*** -.29*** -.32***
Use of drugs or alcohol -.04 -.02 -.18***

* p < .05, two-tailed. ** p < .01, two-tailed. *p<. 001, two-tailed.

 REEVALUATING THE LIFE ORIENTATION TEST 1071

Table 5
Correlations Between (Negative Scale) Factor Scores and Outcomes
Factor

Outcome Optimism Emotional Lability Worry

Depression -.48*** .31*** .05

Intensity of symptoms -.23*** .37*** .26***
Number of symptoms -.23*** .38*** .25***
Active coping .40*** -.06 -.04
Planning .32*** -.03 .05
Suppression of competing activities .14** .08 .07
Restraint coping .17*** -.01 -.17***
Positive reinterpretation and growth .47*** .01 -.02
Use of humor .17*** .10 -.19***
Seeking instrumental social support .25*** .10 .15**
Seeking emotional social support .21*** .12* .20***
Turning to religion .23*** .04 .09
Acceptance .19*** .05 -.15**
Denial -.16** .30*** .11*
Focusing on and venting of emotions -.07 .27*** .40***
Mental disengagement -.15** .42*** .17***
Behavioral disengagement -.42*** .30*** .13*
Use of drugs or alcohol -.12* .24*** .10

* p < .05, two-tailed. **/>< .01,two-tailed. ***p< .001, two-tailed.

significantly with all but one of the 18 outcomes. The sole ex- That is, it correlated significantly with active coping per se, and
ception was the Coping subscale assessing focusing on and vent- with two additional coping tendencies that are often associated
ing emotion. with active coping: planning and the suppression of competing
It is also informative to compare the magnitude and pattern activities. Neither the Emotional Lability factor nor the Worry
of correlations that were obtained for the various factors that factor were significantly related to any of these aspects of active
were extracted in these analyses. With respect to the magnitude coping. In addition, the Optimism factor was the only factor
of the correlations, it is clear that the Optimism factor in each that significantly predicted variations in the use of positive re-
data set predicted about as much unique variance across out- interpretation and growth; Emotional Lability and Worry did
comes as did the remaining two factors that were extracted. In- not.
deed, in both sets of data the average of the absolute values of In closing this section on factor analysis, we should note that
the correlations involving the Optimism factor was higher than the absolute value of the correlation between the Optimism fac-
the averages of the absolute values of the two remaining factors tor scores and raw LOT scores was .89 in each of the two sets
that were extracted (sometimes almost twice as high). of data that were analyzed. This reinforces the notion that the
The data also produced some interesting patterns of associa- Optimism factors were indeed Optimism factors. It also rein-
tions. Consider first the analyses involving the scales assessing forces the idea that the Optimism factors that emerged can be
positive attributes (see Table 4). The most interesting pattern closely approximated by using raw scores on the LOT.
here concerned the set of associations involving social support.
As can be seen, only the Optimism factor predicted a significant Discussion
amount of unique variance in the tendency to seek social sup-
port, and it did this with respect to both facets of support seek- The primary purpose behind these analyses was to reevaluate
ing that were measured. Factor scores from the two remaining the predictive validity of dispositional optimism as assessed by
factors failed to predict either facet. This suggests that the Opti- the LOT. The general conclusion we would like to draw from
mism component of these composite scales may be a more im- the analyses is that the LOT is a viable instrument for assessing
portant correlate of these aspects of social support than either people's generalized sense of optimism. Three sets of findings
the Self-mastery or the Self-esteem component. Of lesser inter- support this conclusion. First, correlations between optimism
est in this data set are the unique associations that emerged be- and the remaining predictors under consideration here were all
tween the Optimism factor and restraint coping, suppression of in the moderate range. Thus, the amount of shared variance
competing activities, use of humor, and the use of religion. Nei- between optimism and the other predictors was also only mod-
ther the Self-mastery nor the Self-esteem factor was significantly erate. Although previous studies have sometimes obtained
related to any of these other outcomes. higher correlations (e.g., Smith et al., 1989), these higher corre-
An interesting pattern of associations also emerged with re- lations have tended to come from studies with smaller sample
spect to the analyses involving the scales that assessed negative sizes. Studies using larger samples have tended to provide values
characteristics (see Table 5). Within this data set, the Optimism closer to the ones obtained here. For example, Mroczek et al.
factor seemed the sole predictor of more active forms of coping. (1993) reported a correlation of —.37 between optimism (as
 1072 M. SCHEIER, C. CARVER, AND M. BRIDGES

measured by the LOT) and neuroticism (as measured by a short with respect to a set of competing predictors. Thinking about
form of the Eysenck Personality Inventory) in a community the third variable problem, however, led us to a more general
sample of 1,192 middle-aged men. Findings such as these and consideration of the conceptual properties of the LOT. More
those reported here suggest that the higher correlations that specifically, we began to reexamine the question of how well the
have sometimes been obtained in the past are best attributed to items on the scale were measuring what they were intended to
perturbations in sampling. be measuring: generalized expectations of good versus bad out-
Second, many of the zero-order correlations between dispo- comes in life.
sitional optimism and the outcomes of interest remained sig- Scheier (1988) has been critical in the past of scales whose
nificant when the influence of the other predictors was partialed meaning is obscured by the presence of items that have little to
out. The partial correlations were strongest for aspects of cop- do with the constructs ostensibly being measured. For example,
ing. These results stand in contrast to those obtained by Smith it is not uncommon to find items such as "I feel blue" on anxi-
et al. (1989). In our view, there are at least two potential reasons ety questionnaires, and items such as "You have usually been
for the discrepancies. First, as just noted, the zero-order corre- optimistic about your future" on scales measuring emotional
lations between optimism and trait anxiety were lower in our instability. Such practices guarantee that such measures of anx-
sample, providing more residual variance to be predicted. Sec- iety will correlate with measures of depression, in the case of
ond, the measure of coping used in our study was more compre- the former example, and that such measures of emotional in-
hensive than that used by Smith et al. (1989). Either or both stability will correlate with measures of optimism, in the case
of these differences may have contributed to the differences in of the latter example. It is obviously better to construct scales in
findings. which such built-in confounds are kept to a minimum. Only by
The results were weakest with respect to reporting of physical minimizing the overlap in item content can accurate estimates
symptoms. In the case of symptom reporting, controlling for be obtained of the degree of correlation between constructs.
other predictors typically eliminated significant associations Examination of the LOT with these considerations in mind
with optimism. Reductions in the magnitude of the associations led us to identify two problematic items, both of which come
were particularly evident and most consistent when trait anxiety from the half of the items that are worded in a positive direction.
and neuroticism were controlled. This aspect of our findings The items are the following: "I always look on the bright side of
closely parallels thefindingsreported previously by Smith et al. things," and "I'm a believer in the idea that 'every cloud has a
(1989), Robbins et al. (1991), and Mroczek et al. (1993). It silver lining.' " What is obvious to us in retrospect is that neither
seems likely that most of the variance in the optimism-symp- item explicitly refers to the expectation of positive outcomes.
tom reporting relationship is due to shared variance with trait Rather, each refers to a particular way of reacting to problems
anxiety, neuroticism, and other conceptually similar variables. and stress—putting problems in the best possible light and
In this regard, it is interesting that trait anxiety and neuroticism searching for hidden benefits and meaning when difficulties
are both variables that are intimately tied to affective experi- arise.
ence. Perhaps it was these ties to affective experience that caused The content of these two items is particularly bothersome, for
trait anxiety and neuroticism to have the effect that they did the following reason. When we first constructed the LOT, we
on symptom reporting (cf. Costa & McCrae, 1985; Watson & were not thinking in terms of identifying mediators of optimism
Pennebaker, 1989). effects. We were simply trying to devise a scale that assessed
A final point involves the factor analyses that were con- optimism as clearly as possible. Since the scale was published,
ducted. Criticisms of the LOT have often ascribed its effects to however, coping has emerged as an important mediator of opti-
positive response bias. It is instructive in this regard that the mism effects (see, e.g., Carver et al., 1993; Scheier & Carver,
factor analyses did not yield a single factor reflecting positivity- 1992). Indeed, one specific, important mediator of optimism
negativity. Rather, when items comprising the various predic- effects is the proclivity to engage in positive reinterpretation and
tors were subjected to factor analysis, a distinct Optimism factor growth. The two problematic LOT items are directed at exactly
emerged in each case, which correlated in the high .80s with this tendency. Thus, any correlation between optimism, as as-
raw LOT scores. Moreover, this Optimism factor predicted a sessed by the LOT, and scales measuring positive reinterpreta-
significant amount of variance in many of the outcomes that tion and growth will be suspect because of the presence on the
were measured. Because the Varimax rotation method used to LOT of two items measuring the same tendencies.
derive the factor scores was based on a principal components The analyses that were reported earlier in this article relied
analysis, the variance attributable to the Optimism factor was on the LOT as it was originally devised. We followed this course
wholly independent of the variance attributable to the other fac- of action because this is the version of the scale that people have
tors that emerged. relied on in the past and are continuing to rely on in their cur-
Taken together, these three considerations clearly suggest that rent and ongoing research. It is also the version of the scale on
the LOT has a good deal of predictive validity, and that disposi- which previously publishedfindingshave been based. For these
tional optimism as assessed by the LOT deserves continued de- reasons, we thought it was important to show that the original
velopment as an independent theoretical construct. version of the scale held up to the challenges it has received.
Because of the concerns we just raised, however, we also went
Study 2: Some Suggestions for Improvement—The back and reanalyzed the data to see whether the associations
Revised Life Orientation Test between optimism and positive reinterpretation and growth
The work reported in this article was prompted primarily by would remain significant if the original eight-item version of
prior research questioning the predictive validity of the LOT the LOT were replaced by a six-item version in which the two
 REEVALUATING THE LIFE ORIENTATION TEST 1073

problematic (coping) items were removed. Both the zero-order Table 6

correlation and all of the partial correlations involving positive
Items Composing the Revised Life Orientation Test
reinterpretation and growth remained significant after the two
items were removed. Removal of the items did cause the corre- Item
lations to drop by about. 10 in value, however, a difference that
was often statistically significant. Thus, the presence of the two 1. In uncertain times, I usually expect the best.
items on the scale does seem to bias the magnitude of the corre- 2. It's easy for me to relax. (Filler item)
lations that are observed.5 3. If something can go wrong for me, it will."
4. I'm always optimistic about my future.
Given the points made in the preceding paragraphs, we be- 5. I enjoy my friends a lot. (Filler item)
lieve it is unwise to continue to incorporate the two items in 6. It's important for me to keep busy. (Filler item)
question when computing scores on the LOT. Elimination of 7. I hardly ever expect things to go my way."
these items, however, leaves only two positively worded items on 8. I don't get upset too easily. (Filler item)
9. I rarely count on good things happening to me.1
the scale. Some researchers consider it advantageous to com- 10. Overall, I expect more good things to happen to me than bad.
pute separate scores for the positively worded and negatively * These items were reverse scored before scoring and analyses.
worded items, either to test the subsets separately (e.g., Marshall
et al., 1992) or to calculate a balance score between positive and
negative items (cf. Michelson, Schwartz, & Marchione, 1991; ticipant's responses to Items 1, 4, and 10 to compute an overall opti-
Schwartz, 1993). Having only two positively worded items mism score. Thus, scores in principle can range from 0 to 24.
makes it difficult to compute separate scores.
These various considerations have led us to begin gathering
Other Scales Administered
data on a revised Life Orientation Test (or LOT-R).6 This re-
vised instrument contains one new positively worded expec- To begin to examine the convergent and discriminant validity of the
tancy item, which increases the number of scorable, positively revised LOT, we asked subjects to complete several additional question-
worded items to three. In addition, we recommend eliminating naires in addition to the LOT-R. A major subset of these additional
questionnaires was composed of the scales used as alternative predictor
one of the negatively worded items from the scoring of the re-
variables in Study 1. More specifically, self-mastery was assessed with
vised instrument, so that equal numbers of positively and nega- Pearlin and Schooler's (1978) Self-Mastery Scale (SMS), trait anxiety
tively worded items are used in the calculation of scores. In the was assessed with Spielberger et al.'s (1974) State-Trait Anxiety Inven-
next section, we describe the administration instructions, tory (STAI), neuroticism was assessed with the Guilford-Zimmerman
scoring procedures, and psychometric properties for the Temperament Survey (GZTS; Guilford et al., 1976), and self-esteem
LOT-R. was measured with Rosenberg's (1965) Self-Esteem Scale (SES). Cron-
bach's alphas for the four scales in the current sample were .75, .88, .86,
and .89, respectively.
Method As an additional measure of neuroticism, a subset of respondents also
completed a 10-item version of the Neuroticism Scale of the Eysenck
Subjects and Procedures Personality Questionnaire (EPQ; Eysenck, 1958; Goh, D. King, & L.
A total of 2,055 undergraduates (622 women, 1,394 men, and 39 who King, 1982). In the 10-item version completed in our research, partici-
did not indicate their gender) from Carnegie Mellon University partici- pants were asked to respond yes or no to statements such as "Do you
pated in the research. Participation was in partial fulfillment of a psy-
chology research requirement. In large group testing sessions conducted 5
across successive semesters from the fall of 1990 to the spring of 1993, This same characterization generally holds for the other analyses
participants completed a number of scales as part of a larger pretesting involving the Life Orientation Test (LOT) that were reported. With few
packet. Participants always completed the revised LOT, but because of exceptions, associations that were significant when using the eight-item
time constraints not all participants completed all of the other scales. scale remained significant when using the corrected six-item scale, but
As a result, sample size varies from analysis to analysis. the correlations involving the six-item scale tended to be slightly smaller
in magnitude. In the case of depression and symptom reporting, differ-
ences between the correlations when using the two versions of the LOT
Scale Format and Administration Instructions for the were nonsignificant. With respect to coping tendencies, 9 of the 14 re-
Revised LOT maining coping dimensions (other than positive reinterpretation and
growth) were significantly less correlated with the altered LOT than they
Items comprising the revised LOT are presented in Table 6. Note that were with the original LOT, three coping dimensions (denial, mental
only 6 of the 10 items on the revised LOT are used to derive an optimism disengagement, and behavioral disengagement) were correlated signifi-
score. Four of the items arefilleritems and are not used in scoring. Of cantly more with the altered LOT than they were with the original scale,
the 6 items that are scored, 3 are keyed in a positive direction and 3 are and two coping dimensions (focusing on and venting of emotions, and
keyed in a negative direction. Respondents are asked to indicate the use of drugs or alcohol) produced no significant differences.
6
extent of their agreement with each of the items, using the following It is unfortunate that data collection on the revised Life Orientation
response format: 0 = strongly disagree, 1 = disagree, 2 = neutral, 3 = Test (LOT-R) only began after the data reported in Study 1 had already
agree, and 4 = strongly agree. Additional instructions caution respon- been collected. As a result, we cannot compare the predictive validity of
dents to be as accurate and honest as they can throughout, and to try the original LOT to the revised LOT with respect to depression, symp-
not to let their answers to one question influence their answers to other toms, and coping. Given the high correlation between the original and
questions. They are explicitly told that there are no right or wrong an- revised LOT, however (in the .90s), we have no reason to believe that the
swers. Negatively worded items (i.e., Items 3,7, and 9) are reverse coded revised scale would have produced any appreciable differences in the
before scoring. Responses to these items are then summed with the par- findings.
 1074 M. SCHEIER, C. CARVER, AND M. BRIDGES

Table 7 five of the six LOT-R items loaded on a single Optimism factor
Factor Loadings for the Revised Life Orientation Test (the lowest loading was .32). The remaining item loaded .330 on
Item
a separate factor but also loaded .327 on the Optimism factor.
Factor 1
In the case of the sample involving the EPQ, a single Optimism
1 .58 factor emerged, which included all six of the items from the
3" .66 LOT-R (the lowest factor loading was .42).
4 .66 In addition to these exploratory factor analyses, the data were
T .79
.74
further examined by confirmatory factor analytic procedures
9" (Joreskog & Sorbom, 1978; Kenny, 1979), using LISREL VI
10 .72
(Joreskog & Sorbom, 1986). Because of the complexity of the
Note. Item numbers refer to items as specified in Table 6. combined item set, confirmatory analyses were limited to the
* Items reversed before analyses. six items comprising the LOT-R. Initially, two simple measure-
ment models were tested, one loading all items on a single factor
and one loading the positively and negatively worded items onto
ever feel 'just miserable' for no good reason?" and "Would you call
separate factors. The models were evaluated by using criteria
yourself a nervous person?" Cronbach's alpha for this sample was .86. described by Bentler and Bonett (1980), Joreskog and Sorbom
(1986), and Bollen (1989). The single-factor solution yielded a
reasonably good fit to the data, x2(9, N = 2,055) = 159.22, p =
Results .001; A, = .95, A2 = .95, root mean square residual (RMR) =
Factor Structure .048, as did the two-factor model, x2(8, N = 2,055) = 43.59, p
= .001, A, = .99, A2 = .99, RMR = .024. Evaluation of the
A principal components factor analysis, using a Varimax final difference in fit between the two models by a hierarchical or
rotation technique, was conducted on a combined sample of nested test (Bentler & Bonett, 1980) suggested that the two-fac-
2,055 undergraduate women and men (see Table 7). The num- tor solution was superior, x2( 1, N = 2,055) = 115.63,/? < .001.
ber of factors retained for final solution was determined by set- Further examination of the covariance matrices from these
ting the eigenvalue criterion at 1.0. The six LOT-R items analyses, however, revealed a pattern of unexpected disturbance
yielded one factor accounting for 48.1% of the variance. As can in the data similar to that reported by Scheier and Carver
be seen in Table 7, all items loaded at least .58 on this factor. (1985). Specifically, there was a higher degree of shared distur-
The mean factor loading was .69. The same one-factor solution bance among positively phrased items than among negatively
also emerged from a subsequent principal components factor phrased items. One explanation is that responses to positively
analysis using an oblique rotation technique. This single factor worded items are more straightforward than responses to nega-
also accounted for 48.1 % of the variance. tively worded items because of differences in semantic complex-
The preceding analyses explored the factor structure of the ity. This may result in greater measurement error due to factors
LOT-R items in isolation. In addition, two principal compo- such as response styles (e.g., yea saying).
nents factor analyses, using Varimax final rotation techniques, As a result of these observations, the simple one- and two-
were conducted on data sets involving two slightly different sub- factor models described above were modified to allow for the
sets of items from all of the scales that were administered. One correlated error among the positively phrased items. The
of these analyses was conducted on a sample of 387 undergrad- models were then re-evaluated. The single-factor solution
uate men and women, and included items from the LOT-R, yielded afitto the data that was substantially better than the one
STAI, SMS, SES, and GZTS. The second analysis was con- that was obtained when the correlated error was not included,
ducted on a sample of 933 undergraduate men and women, and X2(6, N = 2,055) = 16.51, p = .01, A, = .99, A2 = .99, RMR
included items from the LOT-R, STAI, SMS, SES, and EPQ. = .012. The two-factor solution with the same constraints also
The number of factors retained for final rotation in these anal- yielded an acceptable fit, x2(5, N = 2,055) = 16.51, p = .01, A,
yses was determined by setting the eigenvalue at 1.0. In the case = .99, A2 = .99, RMR = .012. In this case, however, the differ-
of the sample involving the GZTS, a single Optimism factor ence in fit between the one- and two-factor models was trivial
emerged, which incorporated five of the six items from the re- and nonsignificant.
vised LOT (the lowest factor loading for thefiveitems was .38).
The sixth item loaded —.439 on a factor defined by itself, but it
also loaded .435 on the Optimism factor. In the case of the sam-
Internal Consistency and Test-Retest Reliability
ple involving the EPQ, a single Optimism factor emerged, Table 8 presents the corrected item-scale correlations for the
which was defined by all six of the items on the LOT-R (the LOT-R in addition to the Cronbach's alpha for the scale if in-
lowest factor loading was .55). dividual items were removed. As can be seen, item-scale corre-
These analyses were subsequently repeated, retainingfivefac- lations ranged from .43 to .63, suggesting that each item is par-
tors for final rotation. We decided to retainfivefactors because tially measuring the same underlying construct, but not to such
the items composing the data set came fromfiveseparate scales. an extent as to be redundant with other items. In addition, all
In addition, scree tests (Cattell, 1966) on the analyses just re- of the items seem to add equivalently to Cronbach's alpha. That
ported suggested the viability of afive-factorsolution. The re- is, the drop in alpha level was relatively comparable as individ-
sults in each case were similar to the results observed in the ual items were removed from the scale (see last column of Table
initial analyses. In the case of the sample involving the GZTS, 8). Cronbach's alpha for the entire six items was .78, suggesting
 REEVALUATING THE LIFE ORIENTATION TEST 1075

Table 8 yses were completed on two independent samples: College stu-

Internal Consistency and Test-Retest Reliability of the Revised dents and patients awaiting coronary artery bypass surgery. The
Life Orientation Test patient group ranged in age from 36 to 82 years (mean = 64.3
years). Thirty-three percent of the group had some education
Corrected item- Alpha with beyond high school. Approximately 80% of the sample was
Item scale correlation item removed married. Table 10 presents the normative data for these two
1 .43 .77 samples. These norms are the only ones currently available for
3" .48 .76 the LOT-R. Norms for other age, class, and occupational
4 .50 .75 groupings will become available as the scale is used on a wider
T .63 .72 basis. Given that the scale has already been successfully admin-
9' .57 .73
.56 .74 istered to two very diverse populations, one community-based
10
and non-college-educated, we do not anticipate difficulty in ad-
ministering the scale to samples with very different
Cronbach's alpha = .78 backgrounds.
Test-retest reliability = .68 (4 months)
= .60 (12 months)
= .56 (24 months) Discussion
= .79 (28 months)
Our intention in removing the two coping items from the
Note. Item numbers refer to items as specified in Table 6. original LOT was to create a scale that focused more exclusively
* Items reversed before analyses. on expectations of good versus bad outcomes. At the same time,
we wanted the psychometric properties of the revised LOT to
remain sound, and to be balanced in terms of positively worded
that the LOT exhibits an acceptable level of internal versus negatively worded items. Only one item was added be-
consistency. cause we wanted to make sure that the items remained centered
One further issue of reliability involves the stability of indi- on expectancies, and it is difficult to write items that are narrow
vidual LOT-R scores over time. To investigate the test-retest in focus but not highly similar or identical to each other. Al-
reliability of the LOT-R, we examined the scores for different though a six-item scale may suffer slight psychometric disad-
samples of college undergraduates who completed the scale at vantages compared with a scale that contains more items, a
two points in time separated by different time intervals. The shorter scale has certain advantages that a longer scale does not.
test-retest intervals for the four groups were 4 months (N = 96), For example, brevity can facilitate use of the scale in research.
12 months (N = 96), 24 months (N = 52), and 28 months (N = Ease of administration can be especially critical in projects that
21). As shown in Table 8, the test-retest correlations were .68, contain long protocols in which subject burden is an important
.60, .56, and .79, respectively. Taken together, thesefindingssug- consideration.
gest that the LOT-R is fairly stable across time.
It is fortunate that the revised LOT does not appear to have
Convergent and Discriminant Validity suffered much, despite its brevity. Internal consistency and test-
retest reliability remain high. In addition, the revised scale
Table 9 presents, separately by gender, the correlations be-
shares only a modest amount of variance with scales measuring
tween the LOT-R and several related scales, including neuroti-
conceptually related concepts—a point we find particularly
cism (measured in two different ways), self-mastery, self-esteem,
noteworthy in light of the criticisms toward which thefirstpart
trait anxiety, and the original LOT. There are several points to
of this article was directed. That is, the modest amount of vari-
be made about the correlations presented in Table 9. First, and
ance shared between the LOT-R and these other constructs
perhaps most important, all of the correlations (except those
should make it easy for future research to disentangle the effects
with the original LOT) are relatively modest in size. The corre-
of optimism from them. Finally, we should mention that we
lations among men ranged from a high of—.52 with trait anxi-
have been able to revise the LOT without distancing ourselves
ety to a low of—.36 with neuroticism as measured by the EPQ.
too far from the original scale on which the revision was based.
The range of correlations among women was similar, ranging
The correlation between the LOT-R and the original LOT is
from a high of .54 with self-esteem to a low of —.36 with neu-
quite high. This high correlation helps ensure continuity of
roticism, again as measured by the EPQ. Second, differences in
findings and understanding as the revised scale slowly replaces
correlations between men and women were negligible. None of
the original instrument as the assessment device of choice.
the six pairs of correlations presented in Table 9 are significantly
different. Finally, the correlation between the revised LOT and
7
the original LOT was quite high for both men and women, sug- For comparison purposes, correlations were also computed between
gesting that the two versions of the scale are assessing highly the original Life Orientation Test (LOT) and the various predictors that
similar characteristics. These high correlations are not surpris- were assessed. The correlations between the original LOT and these pre-
dictors were all higher than the comparable correlations for the revised
ing, given that the two versions of the scale share five items in
LOT that are presented in Table 7. Although the absolute size of the
common.7 differences between correlations tended to be small (ranging from .005
to .037), the differences often proved statistically significant given the
Norms size of the sample involved. Thus, the LOT-R does seem to enjoy a
To investigate norms for men and women, we computed slight comparative edge over the original LOT in terms of the amount
means and standard deviations separately for each gender. Anal- of variance that it shares with related predictors.
 1076 M. SCHEIER, C. CARVER, AND M. BRIDGES

Table 9
Correlations Between Revised Life Orientation Test and Conceptually Related Scales by
Gender
Conceptually related scales Women Men Combined

Original Life Orientation Test

r .95 .95 .95
N 622 1,394 2,055
Self-Mastery Scale
r .46 .51 .48
N 622 1,394 2,055
Trait version of the State-Trait Anxiety
Inventory
r -.54 -.52 -.53
N 616 1,382 2,033
Rosenberg's Self-Esteem Scale
r .54 .50 .50
N 411 982 1,420
Neuroticism assessed by Guilford-Zimmerman
Temperament Survey
r -.49 -.40 -.43
N 323 694 1,041
Neuroticism assessed by items from Eysenck
Personality Questionnaire
r -.36 -.37 -.36
N 293 686 991

Note. All correlations specified here reached significance at p < .001, two-tailed.

The factor structure of the revised LOT deserves special com- LOT-R items in isolation yielded a one-factor solution. One-
ment. As noted earlier, it is not uncommon for the original LOT factor solutions were also reached when factor analyses were
to yield two separate factors, one for positively worded items done on all of the items from all of the related scales simulta-
and one for negatively worded items (e.g., Marshall et al., 1992; neously. Results from the confirmatory factor analyses on the
Scheier & Carver, 1985). Indeed, an exploratory factor analysis six items comprising the LOT-R were more mixed, however.
of the eight items of the original LOT in the data set used in That is, in absolute terms, both the one-factor and the two-fac-
Study 2 again yielded two separate factors. Less clear is the tor model provided an acceptablefitto the observed data. When
proper meaning to be attached to these two separate factors. For correlated error was not included, the two-factor solution
simplicity, we have opted to view optimism and pessimism as proved statistically superior. When models were compared al-
opposite poles of the same dimension, attributing the two-factor lowing for correlated error, the difference between the one- and
structure to differences in item wording rather than more mean- two-factor models proved negligible.
ingful item content (Scheier & Carver, 1985). Others have taken Does this mean that the revised LOT should be scored in a
a different view (Marshall et al., 1992), noting that the optimis- bipolar fashion, using one overall score, or should two separate
tically oriented set of items seems to correlate with a different scores be computed, one for the positively worded items (yield-
group of personality characteristics than does the pessimisti- ing an optimism score) and one for the negatively worded items
cally oriented set of items. (yielding a pessimism score)? Given that the data for the revised
In contrast to analyses involving the original scale, factor scale do provide some justification for treating optimism and
analyses of the revised LOT seem to point in the direction of a pessimism as bipolar, our own approach is to rely on an overall
one-factor structure, but not strongly so. Exploratory analyses score for primary analyses, but to follow up these primary anal-
using both orthogonal and oblique rotation techniques on the yses with subsidiary analyses in which the positively and nega-
tively worded items are examined separately. Although in the
past such separate analyses have generally failed to yield appre-
ciably different results (e.g., Carver et al., 1993), one can imag-
Table 10 ine that in certain contexts the absence of optimism may not
Norms for the Revised Life Orientation Test yield the same results as the presence of pessimism (cf. Marshall
Sample M SD et al., 1992). Although perhaps infrequent, such differential
effects can only be identified by analyzing the positive and neg-
College students 2,055 14.33 4.28 ative items separately.
Women 622 14.42 4.12
Men 1,394 14.28 4.33
Bypass patients 159 15.16 4.05 Concluding Comment
Women 37 14.92 3.97 For any personality construct, there are many alternative
Men 122 15.24 4.09
variables that might be invoked as competitors in the effort to
 REEVALUATING THE LIFE ORIENTATION TEST 1077

predict and understand behavior. In the case now under discus- Bolger, N., & Schilling, E. A. (1991). Personality and the problems of
everyday life: The role of neuroticism in exposure and reactivity to
sion, the variable that may be of greatest interest is neuroticism.
daily stressors. Journal of Personality, 59, 355-386.
Neuroticism holds appeal in part because of its long history as Bollen, K. A. (1989). Structural equations with latent variables. New
a construct in personality psychology, and in part because of its
York: Wiley-Interscience.
place in the list of five "supertraits" that some people believe Carver, C. S. (1989). How should multifaceted personality constructs be
capture most important aspects of personality (e.g., Digman, tested? Issues illustrated by self-monitoring, attributional style, and
1990; John, 1990). The data reported in Study 1, however, re- hardiness. Journal of Personality and Social Psychology, 56,577-585.
vealed unique associations from optimism to depression and to Carver, C. S., Pozo, C , Harris, S. D., Noriega, V., Scheier, M. E, Rob-
several aspects of coping, even after controlling for neuroticism inson, D. S., Ketcham, A. S., Moffat, F. L., Jr., & Clark, K. C. (1993).
(as well as the other predictors that were assessed). These find- How coping mediates the effects of optimism on distress: A study of
ings emerged whether optimism was operationalized with the women with early stage breast cancer. Journal of Personality and So-
original LOT, with an alternative item set in which two prob- cial Psychology, 65, 375-390.
lematic coping items were removed, or with factor scores from Carver, C. S., & Scheier, M. F. (1981). Attention and self-regulation: A
joint analyses of the original LOT and measures of neuroticism control-theory approach to human behavior. New York: Springer-
Verlag.
and trait anxiety. The data join withfindingsof other research-
Carver, C. S., & Scheier, M. F. (1990a). Principles of self-regulation:
ers (e.g., Mroczek et al., 1993) in indicating that optimism and
Action and emotion. In E. T. Higgins & R. M. Sorrentino (Eds.),
neuroticism are not interchangeable. Handbook of motivation and cognition: Foundations of social behav-
As noted in the introduction, however, there is another im- ior (Vol. 2, pp. 3-52). New York: Guilford Press.
portant conceptual issue that stands somewhat behind the Carver, C. S., & Scheier, M. F. (1990b). Origins and functions of positive
scenes here. Neuroticism is usually viewed as a multifaceted and negative affect: A control-process view. Psychological Review, 97,
construct that consists partly, though not entirely, of pessimism 19-35.
(an assumption confirmed in our factor analysis). To ask Carver, C. S., Scheier, M. F., & Weintraub, J. K. (1989). Assessing coping
whether an effect of pessimism is really an effect of neuroticism strategies: A theoretically based approach. Journal of Personality and
(as is sometimes done) begs, in return, a complementary ques- Social Psychology, 56, 267-283.
tion about the effects of neuroticism. Specifically, when exam- Cattell, R. B. (1966). The meaning and strategic use of factor analysis.
ining effects of neuroticism (e.g., Bolger, 1990; Bolger & Schil- In R. B. Cattell (Ed.), Handbook of multivariate experimental psy-
chology(pp. 174-243). Chicago: Rand McNally.
ling, 1991), one might ask whether all facets of neuroticism are
Cohen, S., & Hoberman, H. M. (1983). Positive events and social sup-
important to the effects, or only the part of neuroticism that is
ports as buffers of life change stress. Journal of Applied Social Psy-
related to optimism and pessimism. This question can be ad- chology, 13,99-125.
dressed only if researchers will examine their data in more Costa, P. T., Jr., & McCrae, R. P. (1985). Hypochondriasis, neuroticism,
differentiated ways than has been true in the past. Thus, rather and aging: When are somatic complaints unfounded? American Psy-
than suggest that variables be combined into higher and higher chologist, 40, 19-28.
order, superordinate constructs, which appears to be the pre- Crandall, R. (1973). The measurement of self-esteem and related con-
vailing trend, we suggest that researchers interested in neuroti- cepts. In J. P. Robinson and P. R. Shaver (Eds.), Measures of social
cism decompose their variables into more basic units, such as psychological attitudes (rev. ed., pp. 45-167). Ann Arbor: University
optimism-pessimism, to determine more precisely the exact of Michigan Press.
source of the associations that emerge. Digman, J. M. (1990). Personality structure: Emergence of the five-fac-
tor model. Annual Review of Psychology, 41,417-440.
Eysenck, H. J. (1958). A short questionnaire for the measurement of
References two dimensions of personality. Journal of Applied Psychology, 42, 14-
Aspinwall, L. G., & Taylor, S. E. (1992). Modeling cognitive adaptation: 17.
A longitudinal investigation of the impact of individual differences Friedman, L. C , Nelson, D. V., Baer, P. E., Lane, M., Smith, F. E., &
and coping on college adjustment and performance. Journal of Per- Dworkin, R. J. (1992). The relationship of dispositional optimism,
sonality and Social Psychology, 63, 989-1003. daily life stress, and domestic environment to coping methods used
Beck, A. T, Rial, W. Y., & Rickels, K. (1974). Short form of depression by cancer patients. Journal of Behavioral Medicine, 15, 127-142.
inventory: Cross-validation. Psychological Reports, 34, 1184-1186. Goh, D. S., King, D. W, & King, L. A. (1982). Psychometric evaluation
Beck, A. T., Steer, R. A., & Garbin, M. G. (1988). Psychometric proper- of the Eysenck Personality Questionnaire. Educational and Psycho-
ties of the Beck Depression Inventory: Twenty-five years of evalua- logical Measurement, 42, 297-309.
tion. Clinical Psychology Review, 8, 77-100. Guilford, J. S., Zimmerman, W. S., & Guilford, J. P. (1976). The Guil-
Beck, A. T., Steer, R. A., Kovacs, M., & Garrison, B. (1985). Hopeless- ford-Zimmerman Temperament Survey handbook: Twenty-five years
ness and eventual suicide: A 10-year prospective study of patients of research and application. San Diego, CA: EDITS.
hospitalized with suicidal ideation. American Journal of Psychiatry, John, O. P. (1990). The Big-Five factor taxonomy: Dimensions of per-
142, 559-563. sonality in the natural language and in questionnaires. In L. Pervin
Beck, A. T, Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (Ed.), Handbook of personality theory and research (pp. 66-100).
(1961). An inventory for measuring depression. Archives of General New York: Guilford Press.
Psychiatry, 4, 561-571. Joreskog, K. G., & Sorbom, D. (1978). LISREL: Analysis of linear
Bentler, P. M., & Bonett, D. G. (1980). Significance tests and goodness structural relationships by the method of maximum likelihood. Chi-
of fit in the analysis of covariance structures. Psychological Bulletin, cago: National Educational Resources.
88, 588-606. Joreskog, K. G., & Sorbom, D. (1986). LISREL VI: Analysis of linear
Bolger, N. (1990). Coping as a personality process: A prospective study. structural relationships by maximum likelihood and least square
Journal of Personality and Social Psychology, 59, 525-537. methods. Mooresville, IN: Scientific Software.
1078 M. SCHEIER, C. CARVER, AND M. BRIDGES

Kenny, D. A. (1979). Correlation and causality. New York: Wiley- lation: Translating intention into action. In L. Berkowitz (Ed.), Ad-
Interscience. vances in experimental social psychology (Vol. 21, pp. 303-346). San
Litt, M. D., Tennen, H., Affleck, G., & Klock, S. (1992). Coping and Diego, CA: Academic Press.
cognitive factors in adaptation to in vitro fertilization failure. Journal Scheier, M. F., & Carver, C. S. (1992). Effects of optimism on psycholog-
of Behavioral Medicine, 15, 171-187. ical and physical well-being: Theoretical overview and empirical up-
Marshall, G. N., & Lang, E. L. (1990). Optimism, self-mastery, and date. Cognitive Therapy and Research, 16, 201-228.
symptoms of depression in women. Journal of Personality and Social Scheier, M. F., Matthews, K. A., Owens, J., Magovern, G. J., Sr., Lefeb-
Psychology, 59, 132-139. vre, R. C , Abbott, R. A., & Carver, C. S. (1989). Dispositional opti-
Marshall, G. N., Wortman, C. B., Kusulas, J. W., Hervig, L. K., & Vick- mism and recovery from coronary artery bypass surgery: The bene-
ers, R. R., Jr. (1992). Distinguishing optimism from pessimism: Re- ficial effects on physical and psychological well-being. Journal ofPer-
lations to fundamental dimensions of mood and personality. Journal sonality and Social Psychology, 57, 1024-1040.
of Personality and Social Psychology, 62, 1067-1074. Scheier, M. F., Weintraub, J. K., & Carver, C. S. (1986). Coping with
Michelson, L. K., Schwartz, R. M., & Marchione, K. E. (1991). States stress: Divergent strategies of optimists and pessimists. Journal ofPer-
of mind model: Cognitive balance in the treatment of agoraphobia. sonality and Social Psychology, 51, 1257-1264.
Advances in Behavior Research and Therapy, 13, 193-213. Schwartz, R. M. (1993). The idea of balance and integrative psychother-
Mroczek, D. K., Spiro, A., Ill, Aldwin, C. M., Ozer, D. J., & Bosse, R. apy. Journal of Psychotherapy Integration, 3, 159-181.
(1993). Construct validation of optimism and pessimism in older Smith, T. W., Pope, M. K., Rhodewalt, F , & Poulton, J. L. (1989). Op-
men: Findings from the normative aging study. Health Psychology, timism, neuroticism, coping, and symptom reports: An alternative
12, 406-409. interpretation of the Life Orientation Test. Journal of Personality and
Pearlin, L. I., & Schooler, C. (1978). The structure of coping. Journalof Social Psychology, 56, 640-648.
Health and Social Behavior, 19, 2-21. Spielberger, C. D., Gorsuch, R. L., & Lushene, R. E. (1974). Manual for
Reker, G. X, & Wong, P. T. P. (1983, April). The salutary effects of the State-Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychol-
personal optimism and meaningfulness on the physical and psycho- ogists Press.
logical well-being of the elderly. Paper presented at the 29th Annual Stanton, A. L., & Snider, P. R. (1993). Coping with a breast cancer diag-
Meeting of the Western Gerontological Society, Albuquerque, NM. nosis: A prospective study. Health Psychology, 12, 16-23.
Robbins, A. S., Spence, J. T., & Clark, H. (1991). Psychological deter- Taylor, S. E., Kemeny, M. E., Aspinwall, L. G., Schneider, S. G., Rodri-
guez, R., & Herbert, M. (1992). Optimism, coping, psychological dis-
minants of health and performance: The tangled web of desirable and
tress, and high-risk sexual behavior among men at risk for Acquired
undesirable characteristics. Journal of Personality and Social Psy-
Immunodeficiency Syndrome (AIDS). Journal of Personality and So-
chology, 61, 755-765.
cial Psychology, 63, 460-473.
Rosenberg, M. (1965). Society and the adolescent self-image. Princeton,
Watson, D., & Clark, L. A. (1984). Negative affectivity: The disposition
NJ: Princeton University Press.
to experience aversive emotional states. Psychological Bulletin, 96,
Scheier, M. F. (1988, August). Personality and health: Assessment and
465-490.
implications ofgeneralized outcome expectancies. Paper presented at
Watson, D., & Pennebaker, J. W. (1989). Health complaints, stress, and
the 96th Annual Convention of the American Psychological Associa-
distress: Exploring the central role of negative affectivity. Psychologi-
tion, Atlanta, GA. cal Review, 96, 234-254.
Scheier, M. F , & Carver, C. S. (1985). Optimism, coping, and health:
Assessment and implications of generalized outcome expectancies. Received July 26, 1993
Health Psychology, 4, 219-247. Revision receivedJune 10, 1994
Scheier, M. F., & Carver, C. S. (1988). A model of behavioral self-regu- Accepted July 5, 1994 •

1995 APA Convention Call for Programs

The Callfor Programs for the 1995 APA annual convention appears in the September issue
of the APA Monitor. The 1995 convention will be held in New York, New York, from
August 11 through August 15. The deadline for submission of program and presentation
proposals is December 2,1994. Additional copies of the Call are available from the APA
Convention Office, effective in September. As a reminder, agreement to participate in the
APA convention is now presumed to convey permission for the presentation to be
audiotaped if selected for taping. Any speaker or participant who does not wish his or her
presentation to be audiotaped must notify the person submitting the program either at the
time the invitation is extended or before the December 2 deadline for proposal submission.